Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
MojiDoodle: Handwriting Japanese SR
Learn Japanese with MojiDoodle, a handwriting-based spaced repetition system. Discover its development, the benefits of kinesthetic learning, and future AI integration.
This is work-in-progress system that uses online character recognition (path-based, not OCR) and a system for dynamic segmentation algorithm (messy written answer to list of individual characters) for transcribing the user’s writing into digitized Japanese text. It allows users to write answers to spaced-repetition flashcards using handwriting as opposed to direct keyboard/keypad input.
Ionic/Angular shodo app uses Canvas 2D, SQLite, and Google handwriting recognition.
Japanese learning app collects handwriting data for character segmentation model training.
- LangChainThe open-source framework for building and deploying reliable, data-aware Large Language Model (LLM) applications.LangChain is the essential framework for engineering LLM-powered applications: it simplifies connecting models (like GPT-4 or Claude) to external data, computation, and APIs. The platform provides a modular set of components—Chains, Agents, Tools, and Memory—allowing developers to quickly build complex workflows like Retrieval-Augmented Generation (RAG) pipelines and sophisticated conversational agents. Its Python and JavaScript libraries, combined with LangChain Expression Language (LCEL), offer a standardized interface for rapid prototyping and moving applications to production with confidence.
- OpenAI APIOpenAI API: Your direct gateway to cutting-edge AI models (GPT-4o, DALL-E 3, Whisper), enabling scalable, multimodal intelligence integration into any application.The OpenAI API provides authenticated, programmatic access to a powerful suite of generative AI models. Developers leverage REST endpoints and official libraries (Python, Node.js) to integrate capabilities like advanced text generation (GPT-4o), image creation (DALL-E 3), and speech-to-text transcription (Whisper). This platform is engineered for scale, supporting millions of daily requests for tasks from complex reasoning to real-time customer support agents, ensuring your application gets reliable, state-of-the-art intelligence.
- TypeScriptTypeScript is an open-source superset of JavaScript: it adds static typing and compiles to clean, standards-based JavaScript.TypeScript is a high-level, open-source language developed by Microsoft: it acts as a superset of JavaScript, adding a powerful static type system. This system enables compile-time type checking, catching errors before runtime (a critical benefit for large-scale applications). The TypeScript Compiler (TSC) reliably transpiles all code into clean, standards-based JavaScript (ES3 or newer), ensuring compatibility across any browser or host environment (Node.js, React.js, etc.).
- IonicIonic is the open-source UI toolkit: build high-quality, cross-platform mobile, web, and desktop apps using a single HTML/CSS/JavaScript codebase.Ionic is a powerful, open-source UI toolkit for developing modern, performant cross-platform applications. It allows developers to build for iOS, Android, desktop, and the web from a single codebase, leveraging existing skills in HTML, CSS, and JavaScript. The framework provides a rich library of pre-designed UI components that emulate native app guidelines, focusing on frontend UX and UI interaction. Ionic integrates seamlessly with popular JavaScript frameworks (Angular, React, Vue) and uses tools like Capacitor or Cordova to deploy applications to native app stores, enabling access to native device features (Camera, GPS).
- Google Digital InkGoogle's ML Kit Digital Ink Recognition API converts handwritten strokes into digital text and sketches across 300+ languages using on-device machine learning.This technology tracks touch-screen coordinates (X, Y, and timestamp) to interpret handwriting and drawings in real time. It supports 300+ languages and 1,000+ emojis: all processed locally to ensure speed and data privacy. By utilizing pre-trained models, developers can integrate handwriting recognition into apps for note-taking, math equations, or shape detection without a persistent internet connection.
- Google Input Tools APIIntegrate native-script text input into any application: supporting over 80 languages via Transliteration, IME, and Virtual Keyboard methods.The Google Input Tools API provides core input functionality, enabling users to type in their native script across over 80 languages (e.g., Hindi, Arabic, Chinese). Key features include Transliteration (phonetic conversion, like 'namaste' to 'नमस्ते'), Input Method Editors (IME), and an on-screen Virtual Keyboard. While the original JavaScript API is officially deprecated, developers commonly access the underlying service for real-time, language-agnostic text entry. This service ensures global user communication, offering a seamless input experience for web and mobile platforms.
- Cloudflare PagesCloudflare Pages is a JAMstack platform for frontend developers: deploy fast, secure, and scalable websites directly from your Git repository.Cloudflare Pages delivers a complete CI/CD workflow for your static and dynamic applications. Connect your GitHub or GitLab repository for automatic builds and deployments, complete with shareable preview links for every commit (streamlining collaboration). Your site runs on Cloudflare’s global edge network, ensuring unmatched performance—up to 115% faster than competing platforms—with built-in security (free SSL) and massive scalability. It supports all major frameworks (Next.js, Astro, SvelteKit, etc.) and integrates seamlessly with Cloudflare Workers for adding serverless, dynamic functionality at the edge.
- Cloudflare R2Cloudflare R2: S3-compatible object storage engineered for zero egress fees, eliminating the 'data tax' for developers.R2 is a globally distributed object storage solution built to challenge traditional cloud providers by offering zero egress fees. This core feature ensures you can move data freely, avoiding vendor lock-in and unexpected bandwidth costs. It maintains full Amazon S3 API compatibility, guaranteeing seamless migration and integration with existing tools and workflows. R2 leverages Cloudflare’s global network (330+ data centers) for low-latency access and integrates natively with Cloudflare Workers, allowing you to build high-performance, serverless applications right at the edge.
Related projects
Writing an arXiv Paper (in a Day)
Tokyo
Learn how to leverage AI tools to efficiently write an arXiv paper detailing a novel Japanese/English translation evaluation…
Reality Check: Building a Personal Fact-Checking Framework
Tokyo
Learn to build a personal fact-checking framework using agentic tools to track claims, sources, and argument chains, with…
From Image to Structured Data: Building a Local AI Document OCR Platform for Administrative Workflows
Tokyo
Learn to build a secure, local AI OCR platform for administrative documents. This talk covers practical challenges and…
VibeCoding Workflow Demo
Tokyo
This talk demonstrates a current vibe-coding workflow, covering testing, review, parallel design, development, and supporting tools for dictation…
404 Language Not Found: Building a Tutor from Scratch (and Tears)
Montreal
Building an AI tutor for Tashelhit, a low-resource Amazigh dialect, using Python and LLMs to create lessons from…
MapScroll: Prompt to Maps within minutes
Tokyo
Learn how MapScroll turns simple prompts into interactive, narrative-driven maps, with live demos of creator, educator, and explorer…