Skip to content
View devtechedge's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report devtechedge

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
devtechedge/README.md

Hi, I'm Dev 👋

Full Stack AI Native Engineer · Open Source Contributor

I build production-grade AI and full-stack systems, while contributing fixes upstream to the infrastructure I build on.

My work spans multi-agent systems, RAG, HITL workflows, observability, developer tooling, distributed systems, and Web3 infrastructure.

Python TypeScript Next.js React FastAPI LangGraph PostgreSQL Tailwind CSS Docker Supabase Node.js Claude ChatGPT Grok


🌍 Open Source Engineering

Contribution ledgers

  • 🧩 Non-Web3 OSS Contributions - upstream fixes across AI, developer tooling, databases, runtimes, and JS/Python infrastructure.
  • ⛓️ Web3 OSS Contributions - upstream fixes across wallets, SDKs, blockchain tooling, and Web3 infrastructure.

Merged upstream contributions

Non-Web3

  • pnpm #14754 - fixed pnpm run "/pattern/" --no-bail so one failing matched script no longer cancels its siblings. Non-recursive pattern runs now continue all selected scripts and report the aggregate failure correctly.
  • pnpm #14756 - fixed pnpm update <name>@<version> dropping the dependency's existing ^ or ~ range operator. Updates now preserve the manifest's range style and correctly retain npm: / jsr: prefixes, including prerelease ranges.
  • pnpm #14753 - fixed lockfile: false being ignored when devEngines.packageManager.onFail was download. Automatic package-manager switching now works without creating or updating a project pnpm-lock.yaml, with persistence moved to the global environment when appropriate.
  • Better Auth #11208 - added regression coverage for /phone-number/verify OpenAPI requestBody generation after a Zod intersection had caused the request body to disappear from generated specs. The runtime was unaffected; the fix locks in the expected generated contract.
  • SQLMesh #6040 - fixed a concurrency race in sqlmesh test where ModelTest.create_test() ran on worker threads while time_machine was being started or stopped, causing intermittent IndexError failures. Test creation now happens before work is submitted to the pool.
  • Biome #11667 - added the useBetterDomTraversing nursery lint rule, ported from ESLint Unicorn, to flag unnecessary DOM traversal patterns and provide safe transformations where semantics permit.

Web3

  • thirdweb JS #8938 - fixed useTokenQuery swallowing real token lookup failures and incorrectly turning 401, 429, timeout, and other unexpected errors into unsupported_token. Genuine failures now reach the existing error and retry path.
  • Anza Kit #2032 - corrected the getPatternMatchCodec advanced guide to use number predicates, matching the current codec typing after the related narrowing fix landed.

Selected upstream ecosystems

Non-Web3: pnpm · Biome · SQLMesh · LangGraph.js · Drizzle ORM · CrewAI · node-postgres · TanStack Router · Better Auth · LiveKit Agents · Jinja · Undici · TypeScript-ESLint · Vitest

Web3: Safe · Stellar · Wagmi · RainbowKit · ethers.js · Coinbase Wallet SDK · MetaMask SDK · Reown AppKit · Across Protocol · viem · Solana Web3.js · Ambire


🚀 Flagship Architectures & Projects

  • 🔬 Synthesis - Autonomous multi-agent research: plan → research → synthesize → critique → finalize, with HITL gates, RAG, Reflexion, and live SSE agent graphs (repo).
  • 💼 Jobrow - Live register of still-open US tech roles sourced directly from employer ATS boards, with search, filters, company boards, closed-role tracking, and a public JSON API (repo).
  • ⛓️ Lattice - Live Web3 jobs platform aggregating blockchain and crypto roles from employer ATS boards, with salary observatory, talent directory, gigs, and hiring intelligence (repo).
  • 🪐 Pulsar - Decentralized AI compute protocol interface with Base smart contracts, staking flows, wallet connectivity, 3D visualization, tokenomics, and Foundry-tested contracts (repo).
  • 🧠 AAROP - Explicit Perceive → Plan → Act → Observe → Reflect → Adapt loop with self-verification, bounded autonomy, and replayable traces (repo).
  • ⚖️ RegTrace - HITL Web3 compliance copilot mapping packs onto MiCA/VARA with retrieval-bounded findings and article citations (repo).
  • 🏥 Cadence - Deep-memory healthcare agent lab: multi-layer patient memory, journey stages, consent-scoped clinician briefs (repo).
  • 🔎 Veritas - LangGraph research agent with SSE streaming, Gemini + Tavily when keyed, grounded demo mode otherwise (repo).
  • 🔥 Chaos Simulator - Real-time chaos engineering dashboard with fault injection, self-healing services, animated service topology, scenario orchestration, live telemetry, and recovery analysis (repo).

🛠️ Tech Stack

Category Tooling, Frameworks & Architecture
AI systems & agents Python, LangGraph, LangChain, LangServe, FastAPI, explicit agentic state machines, multi-agent supervisors, HITL interrupts, bounded autonomy, Reflexion / critique loops, tool-use / ReAct, durable checkpoints
RAG, memory & evaluation Hybrid RAG, pgvector, BM25, TF-IDF, JSONB embeddings, cosine retrieval, sentence-transformers, long-term memory, retrieval-bounded generation, eval gates, LLM-as-judge, LangSmith
LLMs, tools & integrations Gemini, OpenAI-compatible providers, Groq, Ollama / local LLMs, Tavily, Telegram Bot API, MCP-oriented tool buses, Google Workspace integrations
Frontend & product TypeScript, React, Next.js, TanStack Start, Vite, Tailwind CSS, shadcn/ui, Lucide, Motion / Framer Motion, Recharts, Three.js, React Three Fiber, HTML5 Canvas, SVG
Data, auth & backend PostgreSQL, Supabase, Neon, PGLite, SQLite, Prisma, Drizzle ORM, SQLAlchemy, Pydantic, Zod, Better Auth, REST, Server Actions, API routes
Realtime, streaming & observability SSE, WebSockets, Socket.io, replayable traces, structured telemetry, OpenTelemetry-shaped tracing, live/demo provider switching
Web3 & smart contracts Solidity, OpenZeppelin, Foundry, Base, viem, wagmi, RainbowKit, ethers.js, wallet SDKs, blockchain / wallet infrastructure
Mobile, runtime & infrastructure Expo / React Native, Node.js, Bun, Express, Docker, Vercel, GitHub Actions, GitHub-hosted automation
Testing & developer tooling Vitest, pytest, Playwright, TypeScript compiler, ESLint / Biome, CI test gates, visual regression, security hardening and threat-model documentation

🎯 Engineering Philosophy

  • Bounded autonomy: explicit state, validation, cost-aware routing, and HITL for irreversible actions.
  • Retrieval-bounded truth: citation-backed RAG with hallucination checks.
  • Observable, upstream, production-ready: inspectable traces, upstream fixes, and tested full-stack systems.
  • Ship the full stack: production-style TypeScript/React surfaces with Python/FastAPI where it fits.

🌐 Connect With Me

LinkedIn X / Twitter GitHub


Pinned Loading

  1. aarop aarop Public

    AAROP is a multi-agent system on an explicit Perceive→Plan→Act→Observe→Reflect→Adapt loop: orchestration, self-verification, resilient recovery, bounded autonomy, replayable traces. Python core wit…

    TypeScript

  2. aegis_vercel aegis_vercel Public

    AEGIS is a multi-agent operations cortex with a live dashboard. Supervisor plus specialist agents (SRE, knowledge, coder, evaluator, communicator), Hybrid RAG, HITL gates, streaming SSE, and a real…

    Python

  3. synthesis synthesis Public

    Autonomous multi-agent research: plan → research → synthesize → critique → finalize. Live agent graph, HITL approval, RAG, Reflexion, streaming SSE, eval gate. Next.js 16, Drizzle, Postgres. Groq l…

    TypeScript 2

  4. healthcare-deep-memory-agents healthcare-deep-memory-agents Public

    Cadence is a pure-Python deep-memory healthcare agent lab: multi-layer patient memory, journey stages, and consent-scoped clinician briefs. No LangChain. Local path is Ollama + SQLite + sentence-tr…

    Python

  5. regulatory_compliance regulatory_compliance Public

    HITL Web3 compliance copilot for VASP licensing reviewers. Maps project packs onto MiCA and VARA with retrieval-bounded findings, article citations, hallucination flags, split-screen accept/edit/re…

    Python

  6. veritas-engine veritas-engine Public

    Self-correcting multi-agent research engine for technical briefs. Planner drafts queries, Tavily retrieves in parallel, a critic grades 1-10 and loops until >=8 or max depth, then a synthesizer wri…

    TypeScript