# AR Labs > AR Labs (Advanced Research Laboratory) is an AI engineering practice founded by Richard M. Thompson, a published cognitive scientist. The lab builds voice agents, AI agent orchestration and the spec-driven engineering methods behind them, and documents its projects, methods and research on this site. Every lab page below is also published as plain markdown in one file: https://arlabs.tech/llms-full.txt. A human-readable map of the whole lab is at https://arlabs.tech/lab. ## Evidence and status - [Evidence and status index](https://arlabs.tech/evidence): What AR Labs has built and tested. Every project with the status its page states (live, building, designing, built, paused or archived), its kind, its last update, and links to its source and related pages. - [Salesbot: a Pipecat voice agent for sales discovery](https://arlabs.tech/projects/salesbot): live · Voice agent · updated 2026-11-13 - [Claude Certified Architect](https://arlabs.tech/projects/claude-certified-architect): live · Certification · updated 2026-11-12 - [arlabs.tech](https://arlabs.tech/projects/arlabs-tech): live · Website · updated 2026-11-12 - [Discovery Copilot for live sales calls](https://arlabs.tech/projects/discovery-copilot): building · Conversation copilot · updated 2026-11-12 - [VOX Presenter](https://arlabs.tech/projects/vox-presenter): building · Voice sessions · updated 2026-11-12 - [Vox Canvas: a browser task graph viewer](https://arlabs.tech/projects/vox-canvas): building · Task graph viewer · updated 2026-11-13 - [VOX Torch: a terminal view over AI agent work](https://arlabs.tech/projects/vox-torch): building · Terminal interface · updated 2026-11-13 - [gc-torch: a terminal view for Gas City](https://arlabs.tech/projects/gc-torch): building · Terminal observability · updated 2026-11-12 - [Taskgraph: a status map for coding agent work](https://arlabs.tech/projects/taskgraph): building · Work status model · updated 2026-11-13 - [Goal Map / Agent Orchestrator: AI agent sessions around the month's goal](https://arlabs.tech/projects/goal-map): building · Agent fleet view · updated 2026-11-13 - [AI agents at AR Labs: the agents we run and what each does](https://arlabs.tech/projects/agents): building · Agent fleet · updated 2026-11-13 - [AI strategy agent: planning reviews from the year down to the day](https://arlabs.tech/projects/strategy-agent): building · Strategy agent · updated 2026-11-13 - [AI marketing manager agent: Night Shift, on the Hermes runtime](https://arlabs.tech/projects/marketing-manager-agent): building · Marketing agent · updated 2026-11-13 - [Vox Prime: a Slack identity that lets coding sessions message AI agents](https://arlabs.tech/projects/vox-prime): building · Agent identity · updated 2026-11-13 - [Resident AI agents: a multi-agent system with roles](https://arlabs.tech/projects/multi-agent-system): building · Multi-agent system · updated 2026-11-13 - [Coding Planner: a spec-driven workflow for Claude Code](https://arlabs.tech/projects/coding-planner): building · Coding pipeline · updated 2026-11-12 · source: https://github.com/richardmthompson/coding-planner - [Cognitive Ecology (book)](https://arlabs.tech/projects/cognitive-ecology-book): building · Book manuscript · updated 2026-11-12 - [VOX Facilitator](https://arlabs.tech/projects/vox-facilitator): designing · Facilitation design · updated 2026-11-12 - [VOX OS: personal AI operating system architecture](https://arlabs.tech/projects/vox-os): designing · Agent orchestration · updated 2026-11-12 - [AI software factory: coding agents with gates](https://arlabs.tech/projects/software-factory): designing · Agent coding pipeline · updated 2026-11-12 - [VoxPlan](https://arlabs.tech/projects/voxplan): built · Android app · updated 2026-11-12 · source: https://github.com/richardmthompson/VoxPlanApp - [TaskGlowTimer](https://arlabs.tech/projects/taskglowtimer): built · Web app · updated 2026-11-12 · source: https://github.com/richardmthompson/TaskGlowTimer - [Break It Down](https://arlabs.tech/projects/breakitdown): built · Claude Code command · updated 2026-11-12 · source: https://github.com/richardmthompson/breakitdown - [Bpai Dûay Gan (Go Together)](https://arlabs.tech/projects/go-together-hackathon): built · Hackathon app · updated 2026-11-12 · source: https://github.com/richardmthompson/bpai-duay-gan - [Voice UI](https://arlabs.tech/projects/voice-ui): paused · Voice interface kit · updated 2026-11-12 - [VOX Scheduler](https://arlabs.tech/projects/vox-scheduler): paused · Calendar assistant · updated 2026-11-12 - [Coherence](https://arlabs.tech/projects/coherence): paused · Voice coaching app · updated 2026-11-12 - [VOX Memory: research into memory for AI agents](https://arlabs.tech/projects/vox-memory): paused · Agent memory research · updated 2026-11-12 - [Vox Manifestor](https://arlabs.tech/projects/vox-manifestor): paused · Voice app · updated 2026-11-12 · source: https://github.com/richardmthompson/voxmanifestor-docs - [Code health skill for Claude Code](https://arlabs.tech/projects/code-health-skill): paused · Claude Code skill · updated 2026-11-12 - [Frontend Designer](https://arlabs.tech/projects/frontend-designer): paused · Design tooling · updated 2026-11-12 - [Second Brain Workshop: build a second brain with Claude Code](https://arlabs.tech/projects/second-brain-workshop): paused · Workshop · updated 2026-11-12 - [AR Labs open-source releases and public repositories](https://arlabs.tech/projects/open-source): no status given ## Projects - [Projects (index)](https://arlabs.tech/projects): What AR Labs has built and is building: voice agents, VOX OS and the AI agent orchestration tools behind it, apps and open-source Claude Code tools. - [Salesbot: a Pipecat voice agent for sales discovery](https://arlabs.tech/projects/salesbot): AR Labs' Salesbot is a browser voice agent built on Pipecat that runs a short sales discovery conversation, reads answers back and closes with a summary. - [Discovery Copilot for live sales calls](https://arlabs.tech/projects/discovery-copilot): AR Labs' Discovery Copilot is a prototype that follows a live sales call, ticks a discovery checklist with quoted evidence and suggests the next question. - [VOX Presenter](https://arlabs.tech/projects/vox-presenter): AR Labs' VOX Presenter runs voice sessions about documents: a fast voice agent discusses a document aloud while a second agent writes analysis on screen. - [Voice UI](https://arlabs.tech/projects/voice-ui): AR Labs' generic browser voice interface and Pipecat backend, taken from the first Salesbot as the start of a reusable voice layer for AI agents. Paused. - [VOX Facilitator](https://arlabs.tech/projects/vox-facilitator): AR Labs' design for a voice facilitator that reads a folder of project documents, then runs a spoken session from understanding to decisions to tasks. - [VOX Scheduler](https://arlabs.tech/projects/vox-scheduler): AR Labs' design for an AI voice calendar assistant for Google Calendar: early prototypes and a documented web design. Paused at the design stage. - [Coherence](https://arlabs.tech/projects/coherence): AR Labs' Android voice-journaling and goal-coaching app: a spoken daily check-in with an AI guide, built as a staged conversation whose state lives in code. - [VOX OS: personal AI operating system architecture](https://arlabs.tech/projects/vox-os): VOX OS is AR Labs' design for an operating layer over AI agent work: one task graph, with terminal and browser views for one person steering many agents. - [Vox Canvas: a browser task graph viewer](https://arlabs.tech/projects/vox-canvas): Vox Canvas is AR Labs' keyboard-driven browser canvas for its task graph: a tree to move through, a panel for each task, and Taskgraph's status map in place. - [VOX Torch: a terminal view over AI agent work](https://arlabs.tech/projects/vox-torch): VOX Torch is AR Labs' terminal interface over its live task graph: portfolio, line of work, single task, what waits on a person, and the agents running. - [gc-torch: a terminal view for Gas City](https://arlabs.tech/projects/gc-torch): gc-torch is AR Labs' read-only terminal view over Gas City, Steve Yegge's agent orchestrator: sessions, work, what waits on a person and an event feed. - [Taskgraph: a status map for coding agent work](https://arlabs.tech/projects/taskgraph): Taskgraph is AR Labs' command-line tool that reads a branch of the task graph with its git and pipeline state and maps where the work stands. - [Goal Map / Agent Orchestrator: AI agent sessions around the month's goal](https://arlabs.tech/projects/goal-map): AR Labs' Goal Map / Agent Orchestrator places every project on a spiral around the month's goal, with the AI agent sessions working on each in orbit. - [AI software factory: coding agents with gates](https://arlabs.tech/projects/software-factory): AR Labs' design for an AI software factory: coding agents take tasks from the task graph to merged code through gates, while a person orders and merges. - [VOX Memory: research into memory for AI agents](https://arlabs.tech/projects/vox-memory): AR Labs' research into long-term memory for AI agents: a framework evaluation, a rejected conversation-memory trial and a paused knowledge-graph design. - [AI agents at AR Labs: the agents we run and what each does](https://arlabs.tech/projects/agents): The AI agents AR Labs runs: a strategy agent, a marketing manager agent, a chat identity for laptop sessions, and coding, research and voice agents. - [AI strategy agent: planning reviews from the year down to the day](https://arlabs.tech/projects/strategy-agent): AR Labs' AI strategy agent runs the lab's planning reviews, yearly to daily, with a person who decides, then records decisions and passes direction down. - [AI marketing manager agent: Night Shift, on the Hermes runtime](https://arlabs.tech/projects/marketing-manager-agent): AR Labs' AI marketing manager agent, Night Shift: it researches, recommends, briefs and measures on a server, while a person writes and approves. - [Vox Prime: a Slack identity that lets coding sessions message AI agents](https://arlabs.tech/projects/vox-prime): Vox Prime is AR Labs' identity in its shared Slack channel for laptop coding sessions, so a session can ask an AI agent for something and read the reply. - [Resident AI agents: a multi-agent system with roles](https://arlabs.tech/projects/multi-agent-system): How AR Labs runs resident AI agents with their own roles, among them a marketing manager and a strategy agent, and how they work with coding agents. - [AR Labs open-source releases and public repositories](https://arlabs.tech/projects/open-source): AR Labs' public code on GitHub: what each repository holds, its state, and the licence GitHub reports for it, including where there is none. - [VoxPlan](https://arlabs.tech/projects/voxplan): AR Labs' open-source Android app for hierarchical goal planning and time tracking, with daily time quotas and a gamified focus timer. - [Vox Manifestor](https://arlabs.tech/projects/vox-manifestor): AR Labs' voice-first Android app in which an AI 'genie' helps a person name a few wishes and map where they are now against where they want to be. - [TaskGlowTimer](https://arlabs.tech/projects/taskglowtimer): AR Labs' keyboard-driven web focus timer that links each task to a goal, rewards every 30 minutes of work, and was the test bed for our design system. - [Break It Down](https://arlabs.tech/projects/breakitdown): AR Labs' open-source Claude Code slash command that turns a long, many-question AI response into a step-by-step conversation, one chunk at a time. - [Coding Planner: a spec-driven workflow for Claude Code](https://arlabs.tech/projects/coding-planner): AR Labs' tooling for its coding pipeline in Claude Code: it routes coding work, plans it before any code is written, and checks each step with scripts. - [Code health skill for Claude Code](https://arlabs.tech/projects/code-health-skill): AR Labs' prototype Claude Code skill for multi-pass code health reviews, based on Steve Yegge's advice on keeping agent-written code healthy. - [Frontend Designer](https://arlabs.tech/projects/frontend-designer): AR Labs' set of Claude Code agents that analyse a live website and extract its visual design into a reviewed prototype and reusable design tokens. - [Second Brain Workshop: build a second brain with Claude Code](https://arlabs.tech/projects/second-brain-workshop): AR Labs' hands-on workshop series teaching coaches, freelancers and small-business owners to build a second brain with Claude Code from plain markdown files. - [Cognitive Ecology (book)](https://arlabs.tech/projects/cognitive-ecology-book): The book manuscript behind AR Labs' cognitive-science approach: Richard M. Thompson on attention, the unconscious, framing and goals, now being restructured. - [Claude Certified Architect](https://arlabs.tech/projects/claude-certified-architect): Richard M. Thompson, who founded AR Labs, holds Anthropic's Claude Certified Architect – Foundations credential. We built a study tracker for the exam. - [Bpai Dûay Gan (Go Together)](https://arlabs.tech/projects/go-together-hackathon): A hackathon app co-built by AR Labs that pairs Thai locals with nomads in Chiang Mai for events and skill swaps, with chat translated by Claude. - [arlabs.tech](https://arlabs.tech/projects/arlabs-tech): AR Labs' public website: a navigable map of the lab as its front page, a reading layer for writing and project pages, and live tools such as a voice agent. ## Concepts - [The ideas AR Labs runs on (index)](https://arlabs.tech/concepts): The vision and the ideas behind AR Labs' work with AI agents, voice interfaces and context engineering, explained for a newcomer. - [The AR Labs vision](https://arlabs.tech/concepts/vision): The AR Labs vision for AI: voice in, concepts out, on an architecture that fits the limits of people and language models, held by individuals and small teams. - [Mission and values](https://arlabs.tech/concepts/mission-and-values): AR Labs builds with machine intelligence for four values: connection and collaboration, education and progression, quality of life for all, sustainability. - [What we believe about AI](https://arlabs.tech/concepts/what-we-believe-about-ai): AR Labs' positions on AI: conversation as a main interface, AI that augments people, people keeping judgement, speed with direction, governed agents. - [The future of work with AI agents](https://arlabs.tech/concepts/the-future-of-work-with-ai-agents): AR Labs' outlook on the future of work with AI agents: agents carrying defined roles while people frame, supervise and judge, and what that asks of people. - [An ecology of mind and machine](https://arlabs.tech/concepts/an-ecology-of-mind-and-machine): The idea behind AR Labs' tagline: people, AI agents and shared records form one system, each mind doing what it is good at, actions meeting a wider check. - [The Loop: how AR Labs works](https://arlabs.tech/concepts/self-developing-self-funding-self-educating): The Loop is the five-stage cycle by which AR Labs learns, experiments, produces, applies and shares, so that each piece of work feeds the next. - [The AI interface architect](https://arlabs.tech/concepts/the-ai-interface-architect): The practitioner AR Labs thinks AI needs: a designer of people-to-agent interfaces, combining agent engineering, voice, visualisation and production skill. - [The lab aesthetic](https://arlabs.tech/concepts/the-lab-aesthetic): AR Labs' visual principles: retro-futurist and vintage-computing inspired, readable first, calm for long thinking, people and AI drawn as collaborators. - [VOX: voice in, concepts out](https://arlabs.tech/concepts/vox): VOX, AR Labs' central idea: speak to AI agents in ordinary conversation and get back a visual map of the ideas and the work, sized for people and models. - [Cognitive ecology for people and AI agents](https://arlabs.tech/concepts/cognitive-ecology): Cognitive ecology, AR Labs' founding framework: small conscious attention frames a large unconscious process, and actions are judged by wider consequences. - [Context engineering for AI agents](https://arlabs.tech/concepts/context-engineering): Context engineering for AI agents as AR Labs practises it: choosing which instructions, knowledge, tools and memory enter a model's context window. - [The harness paradox](https://arlabs.tech/concepts/the-harness-paradox): AR Labs' harness paradox: tightly constrained AI agents do more useful work than loosely configured ones, because each constraint is information. - [Structured conversation for voice agents](https://arlabs.tech/concepts/structured-conversation): AR Labs' structured conversation pattern for voice agents: the language model talks freely while code tracks what has been collected and when to move on. - [Logical levels in agent design](https://arlabs.tech/concepts/logical-levels): Gregory Bateson's logical levels as AR Labs applies them to AI agent instructions: nested frames that each govern the level below, records kept outside. - [The temperature of specification](https://arlabs.tech/concepts/specification-temperature): AR Labs' principle for planning AI coding agent work: start in design-level language and grow specific stage by stage, leaving code detail to later steps. - [The intelligence is in the ordering](https://arlabs.tech/concepts/the-intelligence-is-in-the-ordering): AR Labs' view of AI agent orchestration: every actor, human or agent, needs a task list in the right order, so orchestration is ordering and routing. - [The agentic organisation](https://arlabs.tech/concepts/the-agentic-organisation): AR Labs' design for a company whose roles AI agents can fill: durable threads of work, seats agents occupy, handoffs, status up and direction down. - [The second brain](https://arlabs.tech/concepts/the-second-brain): AR Labs' approach to an AI second brain: your goals, history and preferences in plain files that an AI agent such as Claude Code reads each session. - [Cognitive chunking for AI output](https://arlabs.tech/concepts/cognitive-chunking-for-ai-output): AR Labs' method for presenting AI responses one piece at a time, in chunks sized to working memory, so a long answer becomes a conversation. - [The acceleration paradox](https://arlabs.tech/concepts/the-acceleration-paradox): AR Labs' view of AI productivity: feeling faster is not being faster, and some work, learning above all, should not be sped up. ## Engineering - [How we build software (index)](https://arlabs.tech/engineering): How AR Labs builds software with AI coding agents: spec-driven, test-first work, checks at every step, independent review and a person at the merge. - [How AR Labs engineers software with AI coding agents](https://arlabs.tech/engineering/how-we-engineer-software): How AR Labs builds software with AI coding agents: design first, a spec and tests before code, scripted checks, cross-model review and a person at the merge. - [Software architecture with AI coding agents](https://arlabs.tech/engineering/software-architecture-approach): How AR Labs designs software that AI coding agents build: design before spec, deep modules, C4 diagrams, decision records, risk spikes and tracer-bullet slices. - [Agentic engineering](https://arlabs.tech/engineering/agentic-engineering): What AR Labs means by agentic engineering: building software with and around AI agents by designing harness, context, tools, gates, verification and cost. - [Design principles for AI-generated code](https://arlabs.tech/engineering/design-principles-for-ai-generated-code): Design principles AR Labs applies to code from AI coding agents: build only what was asked, deep modules, two cases before a pattern, one source of truth. - [The coding pipeline for AI coding agents](https://arlabs.tech/engineering/the-pipeline): How AR Labs takes coding work from a rough idea to reviewed, merged code with AI coding agents, a task graph and scripted checks at every step. - [Spec-driven development with AI coding agents](https://arlabs.tech/engineering/spec-driven-development): How AR Labs runs spec-driven development with AI coding agents: a campaign plan (PRP) and a per-task spec grounded in today's code. - [Test-driven development with AI coding agents](https://arlabs.tech/engineering/test-driven-development-with-agents): How AR Labs' AI coding agents work test-first: a failing test, the least code to pass it, tests at deep-module boundaries and a ten-minute regression gate. - [Evidence of completion for AI coding agents](https://arlabs.tech/engineering/evidence-of-completion): How AR Labs gets evidence of completion from AI coding agents: a committed test, check script or screenshot for every criterion, never the agent's word. - [Testing a browser voice agent: what counts as evidence](https://arlabs.tech/engineering/testing-a-browser-voice-agent): How AR Labs tests a browser voice agent: what typed sessions and automated tests can prove, what needs a person listening, and how to state the gaps. - [Code review gates for AI-generated code](https://arlabs.tech/engineering/code-review-gate): How AR Labs gates AI-generated code: a pull request per task, a four-axis review by a second model in a fresh context, and a person's merge. - [Git worktree isolation for parallel AI coding agents](https://arlabs.tech/engineering/worktree-isolation): How AR Labs isolates parallel AI coding agents with git worktrees: each task on its own branch and checkout, and a hook that blocks edits to main. - [Task graphs for AI coding agents with Beads](https://arlabs.tech/engineering/task-graphs-with-beads): How AR Labs keeps a task graph for AI coding agents in Steve Yegge's Beads, and the rules that make the graph safe to dispatch work from. - [Graph engineering for multi-agent systems](https://arlabs.tech/engineering/graph-engineering): Graph engineering for AI agent orchestration as AR Labs applies it: nodes, edges, state, routers and gates, and how our coding pipeline follows them. - [Harness engineering for AI coding agents](https://arlabs.tech/engineering/harness-engineering): Harness engineering at AR Labs: each recurring AI coding agent mistake becomes a hook or check, so rules that matter are enforced, not remembered. - [Implementation walkthroughs](https://arlabs.tech/engineering/implementation-walkthroughs): Why AR Labs' AI coding agents write an implementation walkthrough as they work: a plain account of why each change was made and in what order. - [Reuse before create](https://arlabs.tech/engineering/reuse-before-create): How AR Labs stops AI coding agents duplicating code: search before writing, re-read before each edit, and sweep for the same bug after a race fix. - [Design docs, ADRs and C4 diagrams for AI coding agents](https://arlabs.tech/engineering/architecture-documentation-c4): How AR Labs documents architecture so people and AI coding agents can read it: a design doc, one-page ADRs, and C4 diagrams kept as text. - [Progressive disclosure for AI agent context](https://arlabs.tech/engineering/progressive-disclosure-documentation): How AR Labs layers project documents so AI agents and people load the most constraining context first: outcome, strategy, active work, then tasks. - [Writing rules for AI agents and humans](https://arlabs.tech/engineering/writing-rules-for-agents-and-humans): The plain-language writing rules AR Labs loads into every AI agent session, why grounding matters most, and how a writing-quality skill audits prose. - [Act or ask: when an AI agent should ask permission](https://arlabs.tech/engineering/act-or-ask): How AR Labs' AI agents decide whether to act and report, or stop and ask a person: by reversibility, ownership and whose decision it is. - [Client delivery framework for AI projects](https://arlabs.tech/engineering/client-delivery-framework): The seven-phase process AR Labs uses to take an AI project from first conversation to a running system, with an evidence gate between phases. - [Unattended AI coding agent runs](https://arlabs.tech/engineering/unattended-agent-runs): How AR Labs runs AI coding agents unattended: one execution contract, facts passed in, typed stops, scripted checks and every deviation logged. - [Human-in-the-loop gates for AI coding agents](https://arlabs.tech/engineering/human-in-the-loop-gates): Human-in-the-loop workflow design for AI coding agents at AR Labs: where a person must decide, and how each gate costs one question with a recommendation. - [Model routing and agent economics](https://arlabs.tech/engineering/model-routing-and-agent-economics): How AR Labs routes each AI agent task to a model and effort level, when a subagent costs more than it saves, and why quality evidence, not price, decides. - [Multi-agent orchestration for AI coding agents](https://arlabs.tech/engineering/multi-agent-orchestration): How AR Labs orchestrates AI coding agents on one body of work: a task graph decides what is ready, each task gets a worktree, overlapping work waits its turn. - [Interface contracts between AI agents: who may write what](https://arlabs.tech/engineering/agent-interface-contracts): How AR Labs lets AI agents share state safely: interface contracts drawn between C4 containers, one writer per file, direction down, status up. - [Anatomy of a coding agent harness](https://arlabs.tech/engineering/agent-harness-anatomy): The parts of an AI coding agent harness such as Claude Code (instructions, rules, skills, hooks, subagents, tools, memory) and what each is for at AR Labs. - [Understanding code that AI coding agents write](https://arlabs.tech/engineering/understanding-agent-written-code): How AR Labs keeps people able to read code written by AI coding agents: build walkthroughs, guided readings of a change, codebase tours and a code tutor. ## Skills - [Claude Code skills (index)](https://arlabs.tech/skills): The Claude Code skills AR Labs uses for research, planning, review, writing, diagrams and agent orchestration: what each does and where it came from. - [Coding Planner: Claude Code skill for spec-driven development](https://arlabs.tech/skills/coding-planner): Claude Code skill AR Labs built to route coding work into its spec-driven pipeline: one task, or a campaign that fans out into tracked sub-tasks. - [Grill Me: Claude Code skill for design interviews](https://arlabs.tech/skills/grill-me): Claude Code skill for design interviews, adapted by AR Labs from Matt Pocock's grill-me: one question per decision, then a design brief agents plan from. - [Doc Review: Claude Code skill for reviewing design docs](https://arlabs.tech/skills/doc-review): Claude Code skill from AR Labs for reviewing design docs and specs with cross-model reviewers, then a cold reader who tests whether the redraft can be followed. - [Write Well: Claude Code skill for AI writing quality](https://arlabs.tech/skills/write-well): Claude Code skill for AI writing quality, built at AR Labs: diagnoses AI tells, rewrites and scores prose, and checks first that the reader has the context. - [ResearchBot: evidence-based research for Claude Code](https://arlabs.tech/skills/researchbot): Claude Code research skill from AR Labs: scores each source with the CRAAP test, triangulates major claims, records conflicts, gaps and confidence. - [Multi-agent research: Claude Code skill for web research](https://arlabs.tech/skills/multi-agent-research): Claude Code skill for multi-agent web research, adapted by AR Labs from Daniel Miessler's PAI: 3, 9 or 24 parallel agents, then one sourced report. - [Orchestrate: Claude Code orchestration skill for subagents](https://arlabs.tech/skills/orchestrate): Claude Code orchestration skill from AR Labs: decide whether a task needs subagents, size the fleet, pick each model, brief each agent, then verify. - [To Issues: Claude Code skill to split a plan into tasks](https://arlabs.tech/skills/to-issues): Claude Code skill adapted by AR Labs from Matt Pocock's to-issues: splits a plan into vertical-slice tasks for AI coding agents, wired by dependency. - [Taskgraph: Claude Code skill for AI agent status maps](https://arlabs.tech/skills/taskgraph): Claude Code skill from AR Labs for status maps of AI agent work: what is done and where it landed, what runs next, and which decisions wait for a person. - [Goal map: Claude Code skill to map AI agents onto goals](https://arlabs.tech/skills/goal-map): Claude Code skill from AR Labs that maps AI agent work onto goals: each project on a spiral around the month's goal by priority, its agents as moons. - [C4 diagrams: Claude Code skill for architecture diagrams](https://arlabs.tech/skills/c4-diagram): Claude Code skill from AR Labs for C4 architecture diagrams in Mermaid or Structurizr DSL: fixed colours, labelled arrows, a legend and a review checklist. - [SVG diagrams: Claude Code skill for architecture diagrams](https://arlabs.tech/skills/svg-diagram): Claude Code skill for SVG architecture and agent-flow diagrams, built at AR Labs: hand-positioned, editable text, light and dark themes, one file. - [Paper-and-ink theme: Claude Code skill for design tokens](https://arlabs.tech/skills/paper-ink-theme): Claude Code skill for design tokens, built at AR Labs: one CSS tokens file, a starter page and a written visual language, so agents build pages in one look. - [Comprehend: Claude Code skill to understand AI code changes](https://arlabs.tech/skills/comprehend): Claude Code skill from AR Labs for understanding AI-generated code changes: a guided reading of the diff in runtime order, with notes, a diagram and a quiz. - [Summary: Claude Code skill to summarise AI agent replies](https://arlabs.tech/skills/progressive-summary): Claude Code skill from AR Labs for summarising an AI agent's last reply by progressive zoom: the main point in one sentence, then its parts, strictly nested. - [Break It Down: Claude Code command to chunk long AI replies](https://arlabs.tech/skills/breakitdown): Open-source Claude Code slash command from AR Labs that turns an overwhelming AI reply into one question or chunk at a time, sized to human working memory. - [Design Probe: Claude Code skill for UI design exploration](https://arlabs.tech/skills/design-probe): Claude Code skill from AR Labs for UI design exploration: themed HTML prototypes, desktop and phone screenshots, one change per round, then a locked reference. - [Frontend Implementer: Claude Code skill for responsive UI](https://arlabs.tech/skills/frontend-implementer): Claude Code skill from AR Labs for responsive React and Tailwind UI: finds 13 layout mistakes agents repeat, screenshots five viewports, checks contrast. - [Voice sessions: Claude Code skill for talking to AI agents](https://arlabs.tech/skills/voice-sessions): Claude Code skill from AR Labs that starts a spoken conversation with an AI voice agent about a document, a design interview or a list of items, in the browser. - [Strategy agent skills: planning cadences and escalation](https://arlabs.tech/skills/ceo-agent): Two Claude Code skills behind the AR Labs strategy agent: one runs yearly-to-daily planning cadences as a state machine, one files escalations. - [Dream: Claude Code skill for agent memory consolidation](https://arlabs.tech/skills/dream): Claude Code skill for agent memory consolidation, adapted at AR Labs from OpenClaw's Dreaming: a weekly pass over session transcripts writes a dated record. - [Strategic alignment: Claude Code skill for prioritisation](https://arlabs.tech/skills/strategic-alignment): Claude Code skill for prioritisation, built at AR Labs: when focus breaks down, list active projects, sort them by value and momentum, and pick one thing. - [Advisor mode: Claude Code skill against AI sycophancy](https://arlabs.tech/skills/advisor-mode): Claude Code skill against AI sycophancy, built at AR Labs: the agent challenges your decision, rates its confidence and holds firm without new facts. - [Building in public: Claude Code skill for technical blogging](https://arlabs.tech/skills/building-in-public): Claude Code skill AR Labs uses to turn daily engineering work into technical blog articles, with an eight-part framework and the shorter posts pre-written. - [Create Skill: Claude Code skill for writing agent skills](https://arlabs.tech/skills/create-skill): Claude Code skill for writing agent skills, adapted at AR Labs from Daniel Miessler's PAI: folder structure, trigger descriptions and traceable imports. - [Beads skills: Claude Code skills for the Beads issue tracker](https://arlabs.tech/skills/beads-skills): Two Claude Code skills AR Labs built for the Beads issue tracker: one routes tracker operations through our policy, one links a session to its task. - [Walkthrough: Claude Code skill for codebase walkthroughs](https://arlabs.tech/skills/walkthrough): Claude Code skill for codebase walkthroughs, packaged by AR Labs from Simon Willison's Showboat method: every code excerpt is extracted from the real files. - [Code Tutor: Claude Code commands for learning a codebase](https://arlabs.tech/skills/code-tutor): Claude Code commands from AR Labs for learning code by building it: an agent writes a tutorial on the feature, then tutors you Socratically as you code. ## Research - [Research library: AI agents, voice AI and web platforms (index)](https://arlabs.tech/research): Dated, source-evaluated research reports from AR Labs on voice AI, AI agent orchestration, context engineering, RAG, verification and web platforms. - [AI research agent methodology: how AR Labs evaluates sources](https://arlabs.tech/research/research-methodology): How AR Labs runs research with AI agents: depth levels, CRAAP source scoring, triangulation, recorded conflicts and gaps, and confidence grades. - [Wispr Flow alternatives: local speech-to-text apps for macOS](https://arlabs.tech/research/macos-speech-to-text-alternatives): Which macOS dictation app gives low cost, on-device privacy and system-wide typing? Free and one-time-purchase Whisper-based apps match Wispr Flow. - [Pipecat voice agent architecture: pipeline, STT, TTS and latency](https://arlabs.tech/research/pipecat-voice-agent-framework): How to build a real-time voice agent on Pipecat with Daily WebRTC, Deepgram, ElevenLabs and LangGraph: configuration, latency budgets and known pitfalls. - [Sesame Conversational Speech Model (CSM) and open-source voice AI](https://arlabs.tech/research/sesame-conversational-speech-model): How Sesame's CSM produces natural conversational speech, and which open-source frameworks, speech-to-speech and TTS models can build comparable voice agents. - [Real-time voice AI agent architecture for coding agents such as Claude Code](https://arlabs.tech/research/voice-frontend-architectures): How to put a real-time voice interface in front of a CLI coding agent like Claude Code: multimodal models vs STT-LLM-TTS pipelines, transports, VAD and bridges. - [Building autonomous agents on Hermes: planning, cron and resume](https://arlabs.tech/research/hermes-autonomous-agents): How public builders make Hermes agents plan, run over many sessions and resume. Long runs rely on cron and Kanban, not a plan the agent revises across weeks. - [AI agent planning and execution loop: holding a plan, choosing the next task and resuming](https://arlabs.tech/research/agent-planning-and-execution-loop): How an AI agent should hold a plan, pick its next task among competing goals and resume after a break, drawn from marketing planning and harness design. - [Logical levels and the theory of types applied to AI agent goals](https://arlabs.tech/research/logical-levels-agent-goals): What Bateson's logical levels and Russell's theory of types imply for nesting an AI agent's goals, approaches and tasks, and which failures to avoid. - [AI software factories: a survey of backlog-to-merged-code systems](https://arlabs.tech/research/software-factories-landscape): A survey of more than 60 AI software factories that turn a backlog or spec into verified, merged code. Deploy stages and merge queues are rarest. - [AI agent orchestration driver: reconciliation loops versus events](https://arlabs.tech/research/orchestration-driver-execution-model): Which execution model suits a long-running, crash-prone AI agent orchestration driver? A level-triggered reconciler that re-reads durable state. - [Prompting multi-agent orchestration: delegation rules and failure modes](https://arlabs.tech/research/multi-agent-orchestration-prompting): How should a lead AI agent brief and coordinate parallel worker agents? Anthropic's published practice plus nine recurring failure modes and their fixes. - [Harness engineering for AI agents: origins and sources](https://arlabs.tech/research/harness-engineering-sources): Where did the term harness engineering come from? Hashimoto coined the practice and LangChain the formula Agent = Model + Harness, in early 2026. - [Self-hosted always-on AI agents: OpenClaw, Hermes, Claude Agent SDK](https://arlabs.tech/research/always-on-ai-agents): Self-hosted always-on AI agent platforms in April 2026: OpenClaw, Hermes, NanoBot, NanoClaw, OpenFang and the Claude Agent SDK compared on security and memory. - [Moving AI agent sessions between Claude Code, Codex, Pi and machines](https://arlabs.tech/research/agent-session-portability): How to carry an AI agent session across harnesses and machines. Open-source converters cover much of it, but none moves every native record losslessly. - [AI agent memory for changing facts: append-only records versus memory systems](https://arlabs.tech/research/agent-memory-changing-facts): What memory lets an AI agent know which facts hold now and how they changed. Evidence favours curated, dated records read in full over extracted memory. - [Markdown RAG for Claude Code and LangGraph: grep versus vector search](https://arlabs.tech/research/markdown-rag-coding-agents): How to run RAG over a few markdown files from Claude Code and port it to LangGraph: grep retrieval first, a ChromaDB vector script only for semantic queries. - [RAG frameworks for LangGraph and Claude Code in 2026](https://arlabs.tech/research/rag-frameworks-langgraph): Which RAG frameworks, vector stores, PDF parsers and embeddings suit document Q&A with LangGraph? LangChain plus LangGraph, hybrid search and Qdrant lead. - [Context engineering for long-running AI agents: strategies and memory](https://arlabs.tech/research/context-engineering-long-running-agents): How to keep long-running AI agents effective as context fills: write, select, compress and isolate context, plus memory, checkpointing and AI review. - [Test budgets for AI coding agents: cost proportional to change size](https://arlabs.tech/research/proportional-verification-budgets): How teams scale verification to change size: test impact analysis, staged CI and a ten-minute gate. No AI coding agent found ties tests to change size. - [Why multi-round LLM code review loops diverge](https://arlabs.tech/research/llm-review-loop-divergence): Why repeated LLM code review rounds find more critical issues on smaller problems instead of converging, and which mitigations have the strongest evidence. - [LLM-as-reviewer methods: quality and readability checks](https://arlabs.tech/research/llm-as-reviewer-quality-readability): Which LLM-as-reviewer methods lift quality and readability: fresh-context review, per-criterion rubrics, cross-family judges. Readability scores mislead. - [Displaying a task dependency graph so one person sees what to do next](https://arlabs.tech/research/dependency-graph-display): How to draw a 20 to 60 item work dependency DAG so one person sees what to do next and why. Layered swimlanes with a ready-to-start band came out ahead. - [Scheduling a task dependency graph across AI coding agents](https://arlabs.tech/research/dependency-graph-scheduling): How to run a task dependency graph on a small pool of AI coding agents: scheduling theory favours ready-set list scheduling by critical path over waves. - [Technical blog readability: evidence-based typography and layout](https://arlabs.tech/research/technical-blog-readability): Which typography, contrast, spacing and code-block choices make technical blog posts easiest to read in light and dark mode, graded by strength of evidence. - [Conversational navigation models for LLM voice agents](https://arlabs.tech/research/conversational-navigation-models): How dialogue systems escape rigid linear question flows: agenda and stack navigation, information-state extraction, and LLM tools such as pipecat-flows. - [Node-graph and infinite-canvas tools for a human and AI agent task graph](https://arlabs.tech/research/visual-node-graph-canvas-tools): Which node-graph libraries and infinite-canvas tools suit a shared human and AI agent view of a task dependency graph? React Flow and beads_viewer fit best. - [Cloudflare Workers, D1 and R2 as a default for new web apps](https://arlabs.tech/research/cloudflare-default-platform): Why developers default to Cloudflare Workers, D1 and R2 for new web apps, and which critiques hold: single-region writes, runtime limits, lock-in, no spend cap. - [Technical blog stacks: how Karpathy, Simon Willison, Dan Luu and Julia Evans build theirs](https://arlabs.tech/research/technical-blog-architectures): How seven well-known technical writers build their blogs. Simple, stable infrastructure and plain Markdown go with more posting and easier migrations. - [Vibe coding to vibe engineering: AI-assisted coding in early 2026](https://arlabs.tech/research/vibe-coding-vibe-engineering): How AI-assisted coding moved from casual vibe coding to structured vibe engineering, and the tools, methods and multi-agent patterns of early 2026. - [Writing PRDs and PRPs for AI coding agents: expert methods compared](https://arlabs.tech/research/prd-prp-methodologies): How experts write PRDs and PRPs (product requirement prompts) for AI coding agents: shorter documents, dependency-ordered phases and testable checkpoints. ## Case studies - [Case studies (index)](https://arlabs.tech/case-studies): AR Labs case studies: an anonymised client design engagement, written up as engineering work from discovery to design. - [Music school enquiry and scheduling automation](https://arlabs.tech/case-studies/music-school-automation): An anonymised AR Labs case study: enquiry capture, teacher matching, trial booking and re-placement for a music school, with the owner approving each step. - [Business process discovery for workflow automation](https://arlabs.tech/case-studies/discovery-method-for-workflow-automation): AR Labs' business process discovery method: screen recordings, AI frame analysis and layered evidence show how a small business works before automation. - [Scheduling workspace design for teacher matching](https://arlabs.tech/case-studies/scheduling-workspace-design): AR Labs' interface design for music school teacher matching: a teacher's week beside ranked students, and a rule for which data it owns or mirrors. ## Operations - [AI agent operations: how AR Labs runs (index)](https://arlabs.tech/operations): How AR Labs runs as a small company of one person and a set of AI agents: planning cadences, a strategy agent, a file-based agent inbox and dashboards. - [Strategy agent planning cadences](https://arlabs.tech/operations/strategy-agent-and-cadences): How AR Labs runs a strategy agent: yearly, quarterly, monthly, weekly and daily planning cadences that link long-range direction to daily focus and escalation. - [Running a one-person company with AI agents: departments](https://arlabs.tech/operations/departments-and-activity-streams): How AR Labs organises one person's work into five departments and named activity streams, so a strategy agent can allocate time and agents can take roles. - [AI agent operations dashboards and goal maps](https://arlabs.tech/operations/goal-maps-and-dashboards): AI agent operations dashboard design from AR Labs: what a person must see to supervise agents, and the goal map, week page and cadence dashboard we plan with. - [Monitoring multiple AI agent sessions from one screen](https://arlabs.tech/operations/monitoring-multiple-ai-agent-sessions): What AR Labs thinks a person needs to see and steer many AI agent sessions at once, and the rules we use for returning to a conversation safely. - [Escalation workflow for AI agents](https://arlabs.tech/operations/escalation-between-agents): The escalation workflow AR Labs uses for AI agents: typed, tracked escalations with structured notes, a decision lifecycle and a five-level authority model. - [File-based inbox protocol for AI agents](https://arlabs.tech/operations/agent-inbox-protocol): The file-based inbox protocol AR Labs uses so independent AI agents exchange requests, replies and attachments as durable files, with chat used only to nudge. - [A git repository as a second brain for AI coding agents](https://arlabs.tech/operations/repository-as-second-brain): How AR Labs uses one git repository as memory for AI coding agents: layered documents, one per concern, and one instruction file for Claude Code and Codex. - [Session history and memory for AI coding agents](https://arlabs.tech/operations/session-history-and-memory): How AR Labs gives stateless Claude Code sessions a memory: event logs written by hooks, history search, sessions linked to tasks, weekly consolidation. - [AI scheduling for Google Calendar by domains and modes](https://arlabs.tech/operations/scheduling-by-domains-and-modes): The conversational AI scheduler AR Labs uses to turn weekly priorities into Google Calendar blocks by domains, modes and containers, and to recover lost time. - [How AR Labs demos use data, and their limits](https://arlabs.tech/operations/data-use-and-demo-limits): What the AR Labs Salesbot voice demo collects, where it is stored, which outside services process it, and what the demo is not. A statement of fact. ## Guides - [Guides (index)](https://arlabs.tech/guides): How-to guides from AR Labs for small service business owners who want to fix or automate a workflow, starting with the simplest change before software or AI. - [How to automate music school scheduling from enquiry to booked lesson](https://arlabs.tech/guides/how-to-automate-a-music-school-booking-system): An AR Labs guide for music school owners: capture what a teacher match needs, shortlist teachers, book trial lessons and automate the clerical work. - [How to automate tutor matching and trial lesson booking](https://arlabs.tech/guides/how-to-automate-tutor-matching-and-trial-booking): An AR Labs guide for tutoring companies: turn enquiries and assessments into tutor matches and booked first sessions, and know when your platform is enough. - [What to ask on a music lesson or tutoring enquiry form](https://arlabs.tech/guides/enquiry-form-fields-for-lesson-matching): An AR Labs reference for music schools and tutoring businesses: what to ask on an enquiry form, and why each field matters for matching and trial booking. - [How to turn service call enquiries into dispatch-ready jobs](https://arlabs.tech/guides/how-to-turn-service-call-enquiries-into-dispatch-ready-jobs): An AR Labs guide for plumbing, HVAC and field-service firms: call scripts, forms and software settings so every job is dispatch-ready without call-backs. - [How to automate client intake for an agency or consultancy](https://arlabs.tech/guides/how-to-automate-client-intake-for-an-agency-or-consultancy): An AR Labs guide for agencies and consultancies: turn a vague project enquiry into a confirmed brief before the first call, with forms you may already have. - [How to automate new client onboarding for a bookkeeping practice](https://arlabs.tech/guides/how-to-automate-new-client-onboarding-for-a-bookkeeping-practice): An AR Labs guide for bookkeeping practices: collect a new client's details, documents and software access with one request and reminders, never passwords. - [How to follow up enquiries and quotes without losing them](https://arlabs.tech/guides/how-to-follow-up-enquiries-and-quotes-without-losing-them): An AR Labs guide for small service businesses that lose enquiries and quotes: one list, a next action and owner on each, a routine, then CRM reminders. - [How to map a business process before automating it](https://arlabs.tech/guides/how-to-map-a-business-process-before-automating-it): An AR Labs guide for small business owners: map a workflow as it really runs, count what fails, keep the human decisions, then pick the smallest fix. - [How to collect client facts before a first legal consultation](https://arlabs.tech/guides/how-to-collect-client-facts-before-a-first-legal-consultation): An AR Labs guide for small law firms: collect a prospective client's facts before the first consultation, conflict check first, no advice, a person deciding. - [How to decide what to automate with AI in a small business](https://arlabs.tech/guides/how-to-decide-what-to-automate-with-ai-in-a-small-business): An AR Labs guide to deciding what to automate with AI in a small business: measure costly workflows, pick process, software, rules or AI, test one change. ## Documentation - [Coding Planner documentation](https://arlabs.tech/docs/coding-planner): Full documentation of the AR Labs coding planner: a spec-driven pipeline that carries coding work from an idea to reviewed, landed code with AI coding agents and deterministic checks. - [Coding Planner: The skill](https://arlabs.tech/docs/coding-planner/skill): How the Coding Planner skill routes a coding request as a lone task or a campaign, what each command produces, and where a person's decisions enter. - [Coding Planner: The harness](https://arlabs.tech/docs/coding-planner/harness): How recorded state, evidence checks, approvals and execution hosts carry AI coding work through the coding planner pipeline. - [Coding Planner: Status & next steps](https://arlabs.tech/docs/coding-planner/status): What is built in the coding planner, the remaining gaps in entry, approval and completion, and the deferred capabilities. - [Coding Planner: System reference](https://arlabs.tech/docs/coding-planner/system): The authoritative description of how the coding planner works, in seven parts: concepts, the leaf route, containers, the driver and its hosts, doctrine, what is not built, and configuration. - [Coding Planner: The concepts](https://arlabs.tech/docs/coding-planner/system/concepts): How work is broken up, the two entry doors, the three routes, commands, floors and gates, refuse against escalate, worktrees, and wave and deep modes. - [Coding Planner: The leaf route](https://arlabs.tech/docs/coding-planner/system/leaf-route): The four steps where code is written: generate-spec, implement, review (Gate B) and deliver (Gate C), and what each one proves. - [Coding Planner: Containers](https://arlabs.tech/docs/coding-planner/system/containers): How a container splits work, waits for its children and checks the assembled result, and how the coverage ledger turns each promised scenario into a test. - [Coding Planner: The driver and its hosts](https://arlabs.tech/docs/coding-planner/system/driver-and-hosts): The stateless driver, what it refuses to do, the one-fix review loop, claims and parking, the interactive and headless hosts, and the autonomy dial. - [Coding Planner: Doctrine](https://arlabs.tech/docs/coding-planner/system/doctrine): The standing rules the coding planner rests on: the single writer, guards against floors against gates, the deny primitive, channel discipline, revision-bound approvals and evidence. - [Coding Planner: What is not built](https://arlabs.tech/docs/coding-planner/system/not-built): What the coding planner does not yet do: the end-of-campaign merge, a human-only approval surface, an installation verifier, and the known gaps and risks. - [Coding Planner: Configuration](https://arlabs.tech/docs/coding-planner/system/configuration): Every setting the coding planner reads, where it lives, who writes it, its default and which program reads it: manifest keys, bead metadata, environment variables and labels. - [Coding Planner: The pipeline map](https://arlabs.tech/docs/coding-planner/pipeline-map): Every step and every file in the coding planner pipeline, one region at a time, with a zoomable map of the whole. - [Coding Planner: The entry boundary](https://arlabs.tech/docs/coding-planner/entry-boundary): The handoff out of the grill-me interview and into create-prp, where a deliberately fresh session resumes from recorded state instead of a transcript. - [Coding Planner: Agent reference: coding-planner(7)](https://arlabs.tech/docs/coding-planner/manpage): One screen of load-bearing facts for an agent operating the coding pipeline: routes, steps, floors, transition verbs, labels, reserved metadata and hosts. - [Coding Planner: The two incidents](https://arlabs.tech/docs/coding-planner/incidents): A blameless record of the two planning failures that started the coding planner's rebuild: what happened, the cause, and what changed. - [Goal map: a public snapshot](https://arlabs.tech/goal-map): AR Labs' goal map as it stood on 6 November 2026: the lab's AI agents and the projects they worked on, placed by priority around the month's goal. - [Site map: every page of arlabs.tech as a circle](https://arlabs.tech/canvas): The site drawn by the lab's own tool, VOX Canvas: AR Labs at the centre, its sections around it, then their groups and every page, each one a click away. ## Writing - [The Acceleration Paradox](https://arlabs.tech/blog/the-acceleration-paradox): The AI marketing machine promises 10x acceleration. The research tells a different story. A seven-part series on speed, quality, and what we're actually building. - [What 2 Years of Self-taught AI Engineering actually looks like](https://arlabs.tech/blog/what-2-years-of-self-taught-ai-engineering-actually-looks-like): I have the logs, the version numbers, the line counts, the dead ends and the breakthroughs. This is what 2 years of self-taught AI engineering actually looks like. - [CORE-LOOP: Managing Complex Two-Way Voice Conversation via Agent / LLM](https://arlabs.tech/blog/managing-complex-two-way-conversations-inside-a-core-loop-in-an-android-app): In building voice-enabled AI assistants, one of the most challenging aspects is how to manage the conversation lifecycle. I created Coherence to help users manifest their wishes through structured reflection practices. - [Implementing AI-Assisted Concept Building in an Android Voice Interface](https://arlabs.tech/blog/implementing-ai-assisted-concept-building-in-an-android-voice-interface): I needed to create a system where users could verbally describe concepts related to their goals (or "wishes"), with an AI “brain” driving the conversation flow. - [Designing a Flexible Cognitive Architecture for Human-Computer Interaction](https://arlabs.tech/blog/designing-a-flexible-cognitive-architecture-for-human-computer-interaction): In developing the Vox Manifestor app, I faced a fundamental challenge: how could I create an effective system for representing information exchanged between an AI assistant and a human user? As a cognitive scientist, I knew that the next generation of operating systems needs to work in harmony with how the human brain processes information, not against it. - [STATE-MACHINES: How to Begin Building an Agent from Scratch on Android](https://arlabs.tech/blog/state-machines-how-to-begin-building-an-agent-from-scratch-on-android): Conversational AI: The DNA of the next generation of operating system. In my Android App, Vox_Manifestor, I created a kind of Ai “Genie” who helps the user with 3 high level outcomes they are wishing to achieve. - [Basic Android App Building: Architecture](https://arlabs.tech/blog/basic-android-app-building-architecture): I've been learning Kotlin for Android App development for a few months now. I thought I would make some notes on what I've learned about the basics of app architecture so far. - [One Year Syllabus for Learning Software and A.I. Development "MSc"](https://arlabs.tech/blog/one-year-syllabus-for-learning-software-and-ai-development-msc): 20 years ago (2001-2004) (!!), I studied a Cognitive Science BSc. (Hons) at Exeter University in the UK. So I've developed a year-long syllabus for the self-taught "Master's Programme" I'll be putting myself through. - [Speech Analysis: Accessing whisper.cpp in Python](https://arlabs.tech/blog/speech-analysis-accessing-whispercpp-in-python): In the conclusion of my last article - about my search for a suitable Open-Source Automatic Speech Recognition (ASR) model - I decided to try and pursue whisper.cpp as a model, along with a set of Python bindings that allow us to access the otherwise C++ model with Python. - [Searching for a Python-based Speech Recognition Engine (for CPU Inference)](https://arlabs.tech/blog/searching-for-a-python-based-speech-recognition-engine-for-cpu-inference): To give my Ai learning a context to ground into, I'm writing a funny little app I've called VoxPlan (in Python) which allows you to organise goals and tasks in a hierarchical tree and display them in an interactive GUI. Given its name, the most obvious place to start would be the ability to recognise voice commands. - [Can We Design Intelligent Systems?](https://arlabs.tech/blog/can-we-design-intelligent-systems): The developments in Ai and Machine Learning have advanced to the point I can't ignore them any more, and I've reached the end of a particular trajectory with my most recent career. I'm 40 (well, 41 actually), and I'm starting again. ## About - [About AR Labs and Richard M. Thompson](https://arlabs.tech/about): Who runs the lab, its research focus and how to work with it. - [Curriculum vitae](https://arlabs.tech/cv): Richard M. Thompson's experience, selected work and education. - [Client Delivery Framework](https://arlabs.tech/process): How an engagement with the lab runs, phase by phase. ## Optional - [Full text of every lab page](https://arlabs.tech/llms-full.txt): All lab pages concatenated as markdown, with titles and URLs. - [Sitemap](https://arlabs.tech/sitemap.xml): Every public URL on the site.