# WebSockets Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/websockets
> Markdown URL: https://aitinkerers.org/technologies/websockets.md
> Technology record last updated: 2026-02-24T08:58:51Z
> Generated: 2026-09-22T08:35:39Z

Secures a persistent, full-duplex TCP connection via a single HTTP handshake, delivering low-latency, bidirectional data streaming for real-time applications.

WebSockets establishes a continuous, two-way communication channel: a critical shift from the request/response cycle of HTTP/1.1. The connection initiates with an HTTP `Upgrade` handshake, switching the protocol to `wss` (secure) on port 443. This persistent link minimizes network overhead, eliminating the need for inefficient HTTP polling. The protocol, standardized as IETF RFC 6455 in 2011, ensures near-instant data transfer: vital for high-performance use cases (e.g., live chat, collaborative editing, financial market data).

- Official technology site: https://websocket.org
- Public AI Tinkerers demos and talks: 8
- Result page: 1 of 1

## Recent Public Talks and Demos

### [Orchestrating attention with human-on-the-loop](https://seattle.aitinkerers.org/talks/rsvp_yjRZexQA3q8)

I built a new notification paradigm in https://comment.io, the Multiplayer Markdown Editor, that helps orchestrate not just agents but the people who work with them

- Event context: AI Dev Tools Track - Seattle - July 13 — 2026-07-14 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_yjRZexQA3q8

### [Does your coding harness actually do anything? Measuring it, fairly.](https://boston.aitinkerers.org/talks/rsvp_bx0CiDi1_uo)

Everyone's bolting "harnesses" onto coding agents — Superpowers, GSD, Agent-Skills, Compound Engineering — but does the scaffolding actually make the model build better software, or just feel better? CodingHarness.xyz is an open eval that pits these frameworks head-to-head: same spec, same model, isolated sandboxes, and an evidence-based rubric (does it meet the PRD? is the code any good?). I'll share the surprising result — an inverse-scaling effect lifted straight from a recent Xiaomi paper and reproduced on real runs: harnesses help most exactly where the base model is weakest, and barely at all where it's already strong. Even better, it's framework-dependent — one framework tracks the curve almost perfectly, another inverts it and only helps strong models. Live demo of the eval studio: watch two frameworks build the same app, stream the agent's work turn-by-turn, and see the marginal-gain scorecard light up green/red. Plus a sneak peek at turning framework competitions into a spectator sport (think FIFA bracket, but for coding agents).

- Event context: AI Tinkerers Boston: GTM Agentic AI Launch — 2026-06-29 — Boston
- Public talk page: https://boston.aitinkerers.org/talks/rsvp_bx0CiDi1_uo

### [Hearing the Room Across Language: Building Catch’s Realtime iPhone Interpreter](https://hong-kong.aitinkerers.org/talks/rsvp_vLe9DwZ90cU)

I've been building Catch, a native iPhone app that helps people follow talks, lectures, meetings, panels, classes, and tours in a language they do not fully understand. Catch’s current Listen Mode captures room audio, turns it into live interpreted output, plays interpreted audio through headphones when available, shows transcript text on screen as backup and saves the transcript for review afterward. I initially built Catch for myself so I could better follow conferences and presentations in Chinese. As I kept working on it, I realised that live interpretation can help people join events, talks, meetings, and community spaces they might otherwise avoid because of language barriers. For someone living in another country, or regularly moving between languages, the goal is to make participation feel more possible instead of leaving them reading, guessing, or sitting outside the conversation. Catch differs from products like Google and Apple Translate because it is specifically designed for long-form settings where you need continuous and simultaneous interpretation for extended periods of time.

- Event context: AI Tinkerers Hong Kong & GBA at Dialogue in the Dark (June) — 2026-06-24 — Hong Kong
- Public talk page: https://hong-kong.aitinkerers.org/talks/rsvp_vLe9DwZ90cU

### [GitClaw](https://paris.aitinkerers.org/talks/rsvp_minZf42tgUo)

GitClaw is a real-time voice AI companion designed to help GitHub maintainers manage repositories more efficiently. Built entirely by Nelson PROIA, GitClaw integrates directly into the GitHub workflow and allows maintainers to interact with their repositories through a natural voice interface. Instead of constantly switching between dashboards, notifications, and pull requests, maintainers can simply speak to GitClaw to review PRs, triage issues, track repository activity, and receive real-time updates. By combining voice interaction with intelligent repository awareness, GitClaw reduces the operational overhead of open-source maintenance and allows maintainers to focus on building, reviewing code, and engaging with their communities.

- Event context: High-Performance Local AI Development: Kick-off ThinkStation PGX — 2026-03-17 — Paris
- Public talk page: https://paris.aitinkerers.org/talks/rsvp_minZf42tgUo

### [Orchestrating Claude Inside Vercel Sandboxes for an Isolated Commerce Environment](https://poland.aitinkerers.org/talks/rsvp_-g6_pJg01zE)

This talk walks through how we run Claude inside per-session Vercel Sandboxes to operate on a live Next.js commerce system. Each session provisions an isolated sandbox where Claude connects via the Claude Agent SDK over WebSockets. The initial prompt fetches a Git repository of the storefront, which becomes the working workspace. From there, Claude proposes structured mutations that are applied incrementally as diffs and commits, with optional deployment to Vercel. Commerce state (products, pricing, checkout) is never modified via arbitrary code edits. Instead, mutations flow through typed calls exposed by our Commerce SDK. UI changes can be generative; transactional operations must pass validated primitives. We’re evolving the system toward an in-memory project file representation to reduce cold starts and avoid long sandbox wakeups. Mutations are first applied to this in-memory graph, then executed via Just Bash inside the sandbox, enabling background transitions while keeping the live instance responsive. The demo focuses on the WebSocket orchestration loop, sandbox isolation boundaries, tool schema design, and the failure cases we encountered while ensuring the commerce system remains consistent and deployable at all times.

- Event context: AI Tinkerers Poland x Codex - Meetup in Warsaw #7 (4th March, Wednesday) — 2026-03-04 — Poland
- Public talk page: https://poland.aitinkerers.org/talks/rsvp_-g6_pJg01zE

### [AI Ops Circuit Breaker: Multi-Model Fallback Agent](https://montreal.aitinkerers.org/talks/rsvp_vPSOBmxlZhg)

A live 5-minute demonstration of production AI reliability patterns in action. I'll show how a multi-agent system (Router Agent, Health Monitor, and Fallback Agent) works together to achieve 99%+ uptime even when AI models hallucinate or fail. The demo features: • Real-time hallucination detection using confidence scoring and consistency checks • Circuit Breaker pattern implementation with three states (CLOSED, OPEN, HALF_OPEN) • Automatic failover to a lighter, faster backup model (Gemini Flash) when the primary model (Gemini Pro) becomes unreliable • Live metrics dashboard showing agent communication, system health, and graceful degradation Attendees will watch the circuit breaker open after detecting 3 hallucinations, see the system automatically switch to fallback mode, and observe how the application continues serving requests without downtime. All code is Python + FastAPI with real-time WebSocket updates, and runs 100% locally using Gemini's free API tier.

- Event context: AI Tinkerers Montreal: Demo Night — November 20, 2025 — 2025-11-20 — Montreal
- Public talk page: https://montreal.aitinkerers.org/talks/rsvp_vPSOBmxlZhg

### [CareCompanion AI using conversational AI](https://singapore.aitinkerers.org/talks/rsvp_Y8Vv8oVNXV8)

This project is designed to assist elderly individuals in managing their daily healthcare needs, particularly with remembering and adhering to complex medication schedules and vital health monitoring. At its core is a Virtual AI Caregiver, an intelligent agent that tracks whether the patient has taken all prescribed medications and recorded their vital signs on time. The caregiver maintains a comprehensive history of each patient and records all relevant health data during AI-initiated calls. Over time, this data can be used to generate insightful reports and track health trends. In addition to the caregiver, the system includes a Virtual AI Doctor. If a patient has any medical questions or concerns during a caregiver call—such as clarifications about medications or symptoms—they can request to transfer the call to the virtual doctor. The doctor then provides detailed, personalized explanations and support. The system also allows patients to initiate calls on demand to the virtual caregiver, enabling them to inquire about any aspect of their care plan whenever needed. This seamless integration of monitoring, reporting, and on-demand support offers a comprehensive solution aimed at improving healthcare outcomes and quality of life for elderly patients.

- Event context: AI Tinkerers Singapore: 8th Meetup - Antler x Sambanova - June 20th, 2025 — 2025-06-20 — Singapore
- Public talk page: https://singapore.aitinkerers.org/talks/rsvp_Y8Vv8oVNXV8

### [Build Conversational AI Agents that speak ALL of Singapore’s languages!](https://singapore.aitinkerers.org/talks/rsvp_rYM1OARLGcE)

Singapore is a vibrant tapestry of languages and cultures—and now your AI agents can be too. In this session, discover how to build multilingual conversational agents using ElevenLabs' powerful conversational AI platform. Learn how to seamlessly integrate automatic language detection and real-time switching capabilities to create agents that effortlessly understand and respond in English, Mandarin, Malay, Tamil, and more.

- Event context: AI Tinkerers Singapore: May Meetup - May 21st, 2025 — 2025-05-21 — Singapore
- Public talk page: https://singapore.aitinkerers.org/talks/rsvp_rYM1OARLGcE

## Related Technologies

- [React](https://aitinkerers.org/technologies/react) ([Markdown](https://aitinkerers.org/technologies/react.md)) — 220 public demos
- [Python](https://aitinkerers.org/technologies/python) ([Markdown](https://aitinkerers.org/technologies/python.md)) — 662 public demos
- [TypeScript](https://aitinkerers.org/technologies/typescript) ([Markdown](https://aitinkerers.org/technologies/typescript.md)) — 205 public demos
- [Anthropic API](https://aitinkerers.org/technologies/anthropic-api) ([Markdown](https://aitinkerers.org/technologies/anthropic-api.md)) — 66 public demos
- [AVAudioEngine](https://aitinkerers.org/technologies/avaudioengine) ([Markdown](https://aitinkerers.org/technologies/avaudioengine.md)) — 1 public demo
- [AVAudioSession](https://aitinkerers.org/technologies/avaudiosession) ([Markdown](https://aitinkerers.org/technologies/avaudiosession.md)) — 1 public demo
- [AVFoundation](https://aitinkerers.org/technologies/avfoundation) ([Markdown](https://aitinkerers.org/technologies/avfoundation.md)) — 1 public demo
- [Bun](https://aitinkerers.org/technologies/bun) ([Markdown](https://aitinkerers.org/technologies/bun.md)) — 19 public demos
- [bun:sqlite](https://aitinkerers.org/technologies/bun-sqlite) ([Markdown](https://aitinkerers.org/technologies/bun-sqlite.md)) — 1 public demo
- [Claude Agent SDK](https://aitinkerers.org/technologies/claude-agent-sdk) ([Markdown](https://aitinkerers.org/technologies/claude-agent-sdk.md)) — 16 public demos
- [Claw](https://aitinkerers.org/technologies/claw) ([Markdown](https://aitinkerers.org/technologies/claw.md)) — 1 public demo
- [Cloudflare Durable Objects](https://aitinkerers.org/technologies/cloudflare-durable-objects) ([Markdown](https://aitinkerers.org/technologies/cloudflare-durable-objects.md)) — 2 public demos
- [Cloudflare Workers](https://aitinkerers.org/technologies/cloudflare-workers) ([Markdown](https://aitinkerers.org/technologies/cloudflare-workers.md)) — 22 public demos
- [Docker](https://aitinkerers.org/technologies/docker) ([Markdown](https://aitinkerers.org/technologies/docker.md)) — 147 public demos
- [ElevenLabs](https://aitinkerers.org/technologies/elevenlabs) ([Markdown](https://aitinkerers.org/technologies/elevenlabs.md)) — 42 public demos
- [ElevenLabs Conversational AI](https://aitinkerers.org/technologies/elevenlabs-conversational-ai) ([Markdown](https://aitinkerers.org/technologies/elevenlabs-conversational-ai.md)) — 3 public demos
- [FastAPI](https://aitinkerers.org/technologies/fastapi) ([Markdown](https://aitinkerers.org/technologies/fastapi.md)) — 181 public demos
- [Gemini](https://aitinkerers.org/technologies/gemini) ([Markdown](https://aitinkerers.org/technologies/gemini.md)) — 188 public demos