Riff
Voice agent platform that generates deterministic FSM-based voice agents from plain English business descriptions.
GitHub Portfolio Search
Semantic search engine for personal GitHub repositories using embeddings and TF-IDF.
NVIDIA Parakeet Riva ASR Deployment System
Production-ready deployment pipeline for NVIDIA Parakeet RNNT ASR via Riva with WebSocket streaming and comprehensive infrastructure automation.
Browser Text To Speech TTS Realtime
A client-side, real-time neural text-to-speech engine using WebAssembly and ONNX.
Everyone Ai
Self-service AI workspace platform for launching Claude Code environments via Kubernetes.
COE Overview
Weekly aggregator for security and operational errors from Jira, Wiz, CrowdStrike, and Vibranium.
Generate Title Headline Hooks
An integrity-checked headline generation pipeline that separates proposal, judgment, and verification.
SuperTokens Passwordless Node.js Demo
A prototype implementation of passwordless authentication using SuperTokens magic links, Node.js, and Docker.
Intelligence Briefing Toolkit
A standalone research pipeline that searches the web, extracts claims, accumulates evidence, and generates intelligence briefings via CLI, Library, REST, or MCP.
Picture QR Code Login
Serverless shared photo album with Google sign-in, guest access, and identity-based delete permissions.
Voice Calendar Scheduler FSM
A 24/7 voice-driven apartment viewing scheduling assistant using Twilio/WebRTC, an 8-step FSM, and Google Calendar integration.
Testtestthin
Empty or stub repository with no source code or documentation.
Aws Toolsthin
Empty repository stub for AWS tools.
Codefreshbuildsthin
Thin repository stub for a Codefresh build metrics tool.
PythonArraysthin
An empty or stub repository with no source code or documentation.
Hello World2thin
An empty or minimal GitHub repository with no provided documentation or source code.
k3s App Builder Platform
A self-contained platform for deploying apps with plain language and Claude, running entirely within k3s on a single EC2 instance.
Remote Dev UI
A web-based dual-pane file browser with an integrated terminal for managing local and EC2 files via SSH/SFTP.
AWS Cognito + ALB Auth Portal
A minimal employee access portal using AWS ALB for OAuth2 authentication and FastAPI for group-based authorization.
Real-Time ASR Service with NVIDIA NeMo Parakeet
A production-ready streaming speech-to-text API using NVIDIA's Parakeet-CTC-0.6B model, featuring WebSocket audio ingestion, Silero VAD, and FastAPI.
nano-claw voice loop
A local, voice-powered AI agent that runs in your browser with Metal-accelerated speech recognition and tool execution.
4kUpScalerWorker
Serverless-style image upscaling worker using AWS S3, SQS, and InvokeAI.
AWS_tools_for_pricing
Python scripts to export EC2 instance details and OpenSearch domain configurations for cost analysis.
Audio2ScriptViewer
A Python micro-service that processes transcribed audio messages from AWS SQS and compiles them into a sorted CSV script.
Auth0 to S3 Backend
A reference architecture for mapping Auth0 authenticated users to individual AWS S3 buckets with secure, pre-signed URL access.
FSM-generic (Voice OS)
A generic voice-first operating system framework that defines conversational apps as workflow state machines.
Huberman Lab Podcast Transcripts
A curated collection of MS Word and Markdown transcripts for episodes 1–30 of the Huberman Lab Podcast.
Palindrome-Index
A Python utility that identifies the index of a character whose removal converts a string into a palindrome.
S3 Presigned URL Lambda API Gateway Setup
A Serverless Framework template for generating secure, rate-limited S3 presigned upload URLs via API Gateway and Lambda.
TranscriptionAPI-S3-backend
A lightweight Flask microservice that generates AWS S3 presigned URLs for secure audio file uploads.
UFOBriefing
Static intelligence portal for UAP/UFO disclosure research, powered by the Intelligence Briefing Toolkit.
WhisperLive
A real-time speech-to-text server and client ecosystem using OpenAI's Whisper with WebSocket streaming.
address-validator
FastAPI service that resolves messy/voice-transcribed US addresses to USPS standards via phonetic fuzzy matching and calculates driving routes using OpenStreetMap.
Agent Router
A federated, multi-agent cockpit that spawns Gemini, Claude, and Codex in real PTYs and bridges them to a browser workspace.
art-starry-night-flowfield
WebGPU-accelerated particle flow field visualization inspired by Van Gogh's Starry Night.
Audio Orchestrator
A Python-based service that manages audio file processing workflows by coordinating S3 uploads, database task tracking, and SQS worker queues.
audio-uithin
Browser-based audio recorder with chunking functionality.
audio-ui-cf-s3-lambda-cognito
A deprecated serverless personal cloud platform featuring secure audio recording and file management via AWS Cognito, Lambda, S3, and CloudFront.
audio-ui-realtime-transcribe
Serverless platform for browser-based chunked audio recording, S3 storage, and file management with planned real-time transcription.
audio-workernode
A Python-based distributed worker node for audio transcription tasks, designed to run on RunPod infrastructure and communicate with a central orchestrator.
audio_client_server
A web-based audio recording client that streams sound to a Python Flask server for real-time transcription and processing.
No Prob Bob — Customer Discovery Agent
A conversational AI agent that guides customer discovery interviews, extracts insights, and saves structured seed documents to Afterburner projects.
browser-RAG-retrieval-realtime-night-index-SQLLiteWASM-and-sqllite-vec-portal-vector-db-with-filters
A fully offline, in-browser vector database for real-time RAG retrieval using SQLite WASM and client-side embeddings.
Browser-Native RAG Retrieval Layer
A fully local, WebGPU-accelerated RAG retrieval system that runs embedding and vector search entirely in the browser with zero network calls at query time.
VoiceStream ASR
A production-ready real-time Automatic Speech Recognition service using browser APIs and WebSocket streaming.
browser-llm-local-ai-chat
A zero-build, vanilla JS browser chat interface supporting Ollama, OpenAI, Anthropic, and in-browser WebGPU/WASM inference.
Browser Voice Agent
A fully browser-native voice agent using local WebGPU LLMs, TTS, and STT with no server requirements.
Browser Whisper Models Local Showcase
A client-side web application for private, offline speech-to-text transcription using OpenAI's Whisper models via WebGPU or WebAssembly.
browser_mobile_debug_panel
One-tap debug report button for mobile web apps that captures console logs, browser info, and app state to S3.
cal-provider
Unified async Python library for Google Calendar and CalDAV with an optional MCP server.
Browser-Native RAG Retrieval Layer
A fully local, WebGPU-accelerated RAG retrieval system that runs entirely in the browser with zero network calls at query time.
classVsInstance
Educational Python implementation demonstrating class vs instance variables and basic OOP logic.
claude-chat-workspace
Lightweight self-hosted chat interface for Claude using Node.js, Express, and SSE.
Claude Memory Bridge
A Chrome extension that detects specific memory markers in Claude AI chats and bridges them to external storage via webhooks or serverless functions.
Cloudflare Zero Trust Setup Wizard
Interactive portal and CLI for provisioning secure, identity-gated localhost tunnels via Cloudflare.
codefresh-release-stats
A Python script to count Codefresh Kubernetes releases per cluster over the last week.
CloudDrive: Serverless Personal Cloud with Audio Recording
A secure, mobile-first personal cloud platform featuring chunked audio recording, file management, and event-driven architecture on AWS.
Deploy Portal (Capsule Cloud)
A self-service web portal for provisioning SSH access and deploying applications to an EC2 gateway instance protected by AWS Cognito authentication.
Easy Cognito Nginx Gateway Auth
A production-ready authentication gateway that protects web applications using AWS Cognito, nginx, and oauth2-proxy without requiring code changes.
EventBridge Orchestrator
Infrastructure-as-code repository defining AWS EventBridge schemas, rules, and utility lambdas for a decoupled microservices architecture.
exampleLoops
Educational code samples demonstrating basic loop structures in Java and Python.
flowstate
A declarative Finite State Machine engine for multi-turn conversational AI that uses YAML to define states, typed slots, and per-state model selection.
flowstate-evolve
Analyzes Flowstate workflow traces to recommend graph improvements and generate auditable before/after reports.
GPT-4 & LangChain PDF Chatbot
A Next.js application that enables natural language querying of multiple large PDF documents using GPT-4, LangChain, and Pinecone vector storage.
grassy-knoll
A generic voice-first operating system framework that defines conversational apps as workflow state machines.
Headshot AI Starter
Open-source SaaS starter for generating professional AI headshots using Astria, Next.js, and Supabase.
hello-world
A minimal tutorial repository for Hello World by David Mar.
hello-world-test
A minimal GitHub repository used for testing branch edits and README updates.
html-framework-record-ver2.0
A lightweight, web-only audio recorder that captures PCM data and stores it in IndexedDB for persistent local playback.
ibt-ufothin
Stub repository for a UAP knowledge graph and MetaPath2Vec correlation engine.
interviewnotes
A structured framework and checklist for conducting five-part technical and behavioral engineering interviews.
iPhone & Desktop Companion TTS/SST Talking App
Mac-hosted Python voice assistant streaming WebRTC audio to iPhone with a hybrid FSM workflow engine for complex research.
iPhone Streaming Plus Finite State Machine
Mac-hosted Python voice assistant streaming TTS to iPhone Safari via WebRTC with a hybrid FSM workflow engine for complex queries.
iphone-webrtc-TURN-speaker-streaming-machost-iphonebrowser
Stream generated audio from a Mac host to an iPhone browser via WebRTC and TURN, featuring a voice agent loop with STT, LLM, and TTS.
iran-intel
Static intelligence portal for Iran war investigation signals, powered by the Intelligence Briefing Toolkit.
knightsMovesRecursive
A Python script that calculates valid knight moves on a chessboard using recursive logic and coordinate translation.
Liveblocks Next.js Starter Kit
A production-ready template for building collaborative applications with real-time features using Next.js and Liveblocks.
llm_gpt3.5-vs-gpt4-comparison
A cost-optimized document summarization pipeline using GPT-3.5 that achieves comparable or better results than GPT-4 by leveraging logical paragraph segmentation and hierarchical summarization.
login-system
Serverless authentication backend using AWS Lambda, DynamoDB, and JWT.
mcp-highlighter
A Chrome extension and MCP bridge system that captures LLM-formatted memory blocks from web pages and persists them for AI context.
my-blogthin
Placeholder repository for personal blog writing.
nvidia-RNN-T-Parakeet
A low-latency, GPU-accelerated speech-to-text API server using NVIDIA Parakeet RNN-T and Riva.
nvidia-parakeet
Production-ready deployment system for NVIDIA Parakeet RNNT ASR via Riva with WebSocket streaming.
NVIDIA Riva Conformer-CTC Streaming ASR
Production-ready real-time speech transcription using NVIDIA Riva 2.19 with Conformer-CTC streaming model via WebSocket bridge.
NVIDIA Riva RNN-T Real-Time Transcription System
Ultra-low latency real-time audio transcription system leveraging NVIDIA Riva RNN-T on AWS GPU instances.
NVIDIA RNN-T Production Transcription System
Automated deployment of a GPU-accelerated SpeechBrain Conformer RNN-T speech recognition server with WebSocket streaming and S3 integration.
openai_transcribe
A lightweight Node.js HTTP server that proxies audio files to the OpenAI Whisper API for transcription.
Open Source Sound Generator LLMs
A local browser-based benchmarking suite for comparing nine open-source audio generation models with real-time progress streaming and Apple Silicon optimization.
phone-agent-keiko-tv
A Python-based AI phone agent for keiko.tv that handles customer service calls, manages contacts via SQLite, and integrates with FSM-generic for state-driven scheduling.
phone-agent-scheduler
A real-time voice AI agent for trades businesses that handles phone calls, checks Google Calendar availability, and books appointments using a deterministic FSM and LLM safety nets.
Picture QR Code
A serverless, password-gated shared photo album accessed via QR code, using AWS Lambda and S3.
pictures-valentina
Automated pipeline for generating a 16-year school portrait poster from an Apple Photos library.
pythonArrays2
A Python script that reads an array of integers from standard input and prints them in reverse order.
rag-document-chat
A Python-based Retrieval-Augmented Generation (RAG) system for document Q&A using FastAPI, Streamlit, ChromaDB, and OpenAI.
RAG Document Chat System v2
A modular, production-ready Retrieval-Augmented Generation system with hierarchical document processing and paragraph-aware search.
recording_app
Browser-based voice memo recorder with chunked S3 streaming and local faster-whisper transcription.
riot
A lightweight Python socket server and curses-based client for tracking taxi driver performance ratings.
search-tool-provider
Unified async web search interface supporting 6+ backends with MCP server, CLI, and admin UI.
sip-voice-transport
PSTN phone call transport for Mac-local voice AI supporting Telnyx and Twilio.
Smart Transcription Router
Hybrid AWS architecture routing audio transcription between real-time GPU FastAPI and batch SQS processing based on server health.
spa_react_javascript_auth0
Auth0 React SPA authentication sample with auxiliary S3 audio upload utilities.
Iris Kade: Local Browser-Native Conversational AI
A zero-network, WebGPU-accelerated conversational AI with dual-LLM architecture, multi-lane RAG, and streaming TTS/STT running entirely in the browser.
Iris Kade v2: Local Browser AI with Intent Workflow
A zero-network, WebGPU-powered conversational AI featuring an embedder-based intent classifier and a visual workflow editor for stateful dialogue.
Speaker Workflow System
A visual editor and runtime engine for declarative, multi-turn conversational workflows defined in JSONL.
spot-setupthin
A thin repository intended for setting up GPU spot instances optimized for running Whisper.
SSH Helper
A lightweight, web-based SSH terminal emulator with IP whitelisting and external authentication support.
stringsAndLoops
A Python script that separates even and odd indexed characters from input strings.
stripe-single-subscription-ec2
A Python Flask backend and vanilla JS frontend implementing Stripe Checkout for single subscriptions, configured for EC2 deployment.
SuperTokens S3 Example
A reference implementation of passwordless authentication using SuperTokens, Go, and Docker on AWS EC2.
switchboard-ai
Append-only coordination channel for AI agents and humans via HTTP and JSONL.
FriendNet
A private network dashboard that aggregates mDNS service discovery from multiple Tailscale-connected homes into a single view.
Persona RAG Engine
A multi-domain intelligence platform that isolates analytical personas and evidence corpora to prevent data bleed while enabling specialized, voice-consistent Q&A.
tool-browserAutomationScripts
macOS CLI and library for listing Chrome profiles, decrypting cookies, and launching browser instances.
tool-messaging-projectthin
Empty or stub repository for a messaging tool project.
tool-s3-cloudfront-push
A shell-based static site deployer that crawls a running local website and pushes it to AWS S3 and CloudFront for low-cost, HTTPS-enabled global hosting.
Video Picture Annotator Review
A Playwright-style tool for analyzing screen recordings, annotating UI elements with AI, and generating visual diff reports.
whatsup
Multi-transport project messaging for Afterburner sprints via Telegram, WhatsApp (planned), and console.
Traceable Searchable ADR Memory Index
A framework for traceable, searchable project memory with multi-agent sprint automation and architecture decision tracking.
Phone Agent Training Pipeline
A pipeline for generating synthetic phone conversation data and fine-tuning local LLMs (via LoRA on Apple Silicon) to act as automated plumbing receptionists.
transcriber-2-pass-riva-conformer-cf-s3-lambda-cognito-adapter-2025-10-14
Hybrid architecture combining NVIDIA Riva Conformer-CTC for real-time streaming ASR with an AWS serverless backend for secure, chunked audio storage and session management.
transcription-end-to-end-docs
Documentation and integration guide for a serverless transcription pipeline using AWS EventBridge, Cognito, and Lambda.
transcription-realtime-whisper
Production-ready real-time speech transcription system supporting WhisperLive edge proxy and NVIDIA Riva enterprise architectures.
CloudDrive: Real-Time Transcription Platform
A production-ready SaaS for real-time speech-to-text using WhisperLive, AWS Cognito, and serverless backend.
transcription-sqs-spot-s3
Scalable AWS-based audio transcription system using SQS queues, EC2 Spot instances, and WhisperX/Voxtral models.
ui-audiorecorder-html-sample
Reference implementation for recording independent, playable audio chunks in the browser to avoid WebM container header issues.
voice-frontend-modules
Composable Python infrastructure for voice AI apps: WebRTC transport, edge authentication, and pluggable STT/TTS/LLM engines.
Voice Base — Browser-to-Server Voice UI
A minimal, self-contained voice conversation system connecting browser microphone to Claude via WebRTC, using local Whisper for STT and Piper for TTS.
voice-optimal-RAG
Self-contained Dockerized RAG service for semantic document search using LanceDB and Nomic embeddings.
Voice Print
Real-time speaker diarization and voice fingerprinting system using multiple embedding models.
website-cloner
A six-phase static site cloning tool that enumerates URLs, downloads assets, rewrites links, detects dynamic content, and deploys to AWS S3.
whisper-runpod
Containerized Faster Whisper transcription service optimized for RunPod with S3 integration.
whisperX-runpod
Scripts and handlers for deploying WhisperX with speaker diarization on RunPod or AWS EC2 GPU instances.
whisperlive-runpod
Deployable real-time speech-to-text service using OpenAI Whisper on RunPod GPU infrastructure.
whisperlive-salad
Real-time speech-to-text server using OpenAI Whisper, optimized for Salad Cloud GPU deployment.
youtube_commercial_detector
A utility that estimates YouTube ad frequency by downloading videos and analyzing their text transcriptions for commercial keywords.
YouTube Phrase Scanner
A serverless, GPU-accelerated pipeline that downloads YouTube videos, transcribes them with WhisperX, and scans for specific phrases.
YouTube Phrase Scanner
Serverless pipeline for downloading, transcribing, and scanning YouTube videos for specific phrases using WhisperX and AWS infrastructure.