Skip to content
View NoumanZahid-85's full-sized avatar
🎯
Focusing
🎯
Focusing
11 followers · 18 following

Achievements

Achievement: Pair ExtraordinaireAchievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Achievements

Achievement: Pair ExtraordinaireAchievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Highlights

  • Pro

Block or report NoumanZahid-85

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
NoumanZahid-85/README.md

typing identity

Portfolio Website  LinkedIn  Email  Status

developer animation

I design and ship production-grade AI systems: real-time voice agents, self-reflective RAG pipelines,
and cloud-deployed ML, where the LLM stays a language interface and deterministic engineering does the heavy lifting.

🎙️ Real-Time Voice AI

• Sub-500ms TTFT full-duplex conversational voice agents
• Pipeline: Silero VAD · Whisper STT · LangGraph · LiveKit / Orpheus
• Protocols: Pure WebSockets, low-latency streaming & dynamic barge-in

🧠 Agentic RAG & Graph Workflows

• Self-Reflective RAG: Query rewriting, hallucination grading & abstention
• Vector Engines: pgvector, Qdrant, multi-hop semantic graph routing
• Benchmarking: RAGAS evaluation across 200k+ document biomedical corpuses

⚙️ Deterministic Engineering

• Zero Race Conditions: PostgreSQL row-level locks for concurrent booking
• State Machines: LangGraph state graphs with strict Pydantic validation
• Resilience: Redis distributed state caching, retry circuits & rate limiting

☁️ Cloud Architecture & MLOps

• Containerization: Dockerized microservices & Kubernetes orchestration
• AWS Ecosystem: S3, SageMaker, EC2, ECS, Lambda, IAM & VPC
• CI/CD: Automated linting, test suites & deployment via GitHub Actions

divider

Tech Stack & Ecosystem

Languages languages
AI · ML · LLMs ai-ml
Backend · DB backend-db
Cloud · DevOps cloud-devops
Frontend · Tools tools

divider

Featured Engineering Projects

[ 01 ] OmniVoice AI · Voice Commerce Agent

Low-latency conversational voice agent that answers catalog questions and executes checkout on a live call: pure WebSocket pipeline (Silero VAD → Whisper STT → LangGraph → Orpheus TTS), RAG-grounded responses, sub-500ms first response.

live demo source

Python FastAPI Next.js LangGraph Groq Qdrant PostgreSQL Redis

[ 02 ] Tradecraft HR · Grounded AI Recruiter

Autonomous recruiter platform that parses a candidate's real GitHub repositories to dynamically generate deep technical interview questions, runs a lip-synced avatar interview with barge-in, and issues scorecard reports backed by transcript evidence.

live demo source

Python FastAPI Next.js LangGraph Groq Whisper SQLite ReportLab

[ 03 ] Self-RAG · Biomedical Research Assistant

Self-reflective RAG over a 211k-document PubMed corpus: LLM-judged relevance grading, hallucination checking, iterative query rewriting, and principled abstention when corpus evidence runs out.

source

Python FastAPI LangGraph Groq PostgreSQL pgvector React RAGAS Docker

[ 04 ] Car Wash Dual-Channel Booking Agent

WhatsApp (Baileys + Groq) and live-voice (LiveKit + Uplift AI) agents sharing one deterministic FastAPI backend: Postgres row-level locking makes double-booking mathematically impossible across channels.

source

Python Node.js FastAPI Baileys LiveKit Groq Supabase PostgreSQL

[ 05 ] Compakt · Smart URL Shortener

High-throughput link shortening suite featuring password-protected links, TTL expiring redirects, burn-after-reading URLs, dynamic QR code generation, and click analytics powered by Supabase.

source

React TypeScript Supabase Vercel JWT

[ 06 ] PrintWall · Instant Film Social Feed

Full-stack photo-sharing feed where uploads convert to instant-film aesthetics: JWT authentication via fastapi-users, drag-and-drop media streamed to ImageKit CDN, on-the-fly WebP transforms for fast loads.

source

Next.js React TypeScript Tailwind FastAPI ImageKit SQLite

divider

GitHub Stats & Activity

GitHub stats streak stats

top languages

divider

Let's Connect & Collaborate

Portfolio Website   LinkedIn   GitHub   Email


footer

Designed for high-throughput AI engineering · Always open to innovative projects and tech discussions.

Pinned Loading

  1. speed-to-lead speed-to-lead Public

    AI lead router - enriches, scores, and routes inbound leads to Slack (hot) or email (warm) in under 10 seconds. FastAPI + LangGraph + Groq + Neon Postgres, running on a /month free-tier stack.

    Python

  2. Self-RAG-Bio-Medical-Research-Assistant Self-RAG-Bio-Medical-Research-Assistant Public

    An agentic Self-RAG system that answers biomedical research-verification questions using a LangGraph pipeline — retrieves from PubMed abstracts, grades its own retrieval, checks for hallucination, …

    Python

  3. clinic-whatsapp-ai-agent clinic-whatsapp-ai-agent Public

    WhatsApp chatbot AI Agent for CityCare Clinic - built with Baileys, LangChain, RAG (ChromaDB + HuggingFace), two-tier memory, and LangGraph agent. Polyglot microservices architecture: Node.js gatew…

    Python

  4. car-wash-booking-agent car-wash-booking-agent Public

    Dual-channel conversational booking agent (WhatsApp + Voice) for a single-location car wash. Both channels share one deterministic FastAPI booking engine backed by PostgreSQL row-level locking (SEL…

    HTML