Skip to content
View Rayyan9477's full-sized avatar
💥
Building
💥
Building

Highlights

  • Pro

Block or report Rayyan9477

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.

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
Rayyan9477/README.md
Rayyan Ahmed - Founding AI Engineer building production AI systems from scratch

Email LinkedIn Resume 2026 Portfolio

Founding AI Engineer Production AI RAG and LLM healthcare workflows Open to collaboration

Profile Views
2,259
Profile views
Followers
78
Followers
Total Stars
164
Total stars
Current Streak
308
Day streak

Focus

I am a founding AI engineer who builds the first version of things that do not exist yet: AI infrastructure, agent workflows, retrieval systems, validation loops, and production automations. My work turns messy operational processes into reliable pipelines that teams can actually use.

Agentic AI

Multi-agent LangGraph workflows, browser automation, MCP servers, tool orchestration, and guardrail-driven execution.

Document Intelligence

OCR/VLM extraction, adaptive schemas, duplicate detection, validation loops, and structured exports for large batches.

RAG and MLOps

Hybrid retrieval, LLM integrations, drift monitoring, retraining pipelines, API serving, and production observability.

Current Work

Founding AI Engineer at Nobility RCM | Islamabad, Pakistan | March 2025 - Present

Joined as the company's founding AI engineer and started the AI function from scratch: technical direction, architecture, model workflows, automation strategy, evaluation patterns, and production delivery.

  • Established the initial AI engineering stack for healthcare operations, including LangGraph agents, RAG pipelines, MCP integrations, Playwright browser automation, and validation-first LLM workflows.
  • Built a medical compliance agent that reads clinical notes and extracts ICD-10, CPT, and HCPCS billing codes with a fine-tuned Phi-3 model, a 4-agent LangGraph workflow, and layered anti-hallucination checks.
  • Developed LLM-powered browser automation agents with Playwright and Browser Use to navigate billing portals, extract claim data, and auto-fill repetitive workflows.
  • Designed retrieval and knowledge workflows with FAISS vectors, BM25 keyword search, and internal system integrations so teams can query billing data in natural language.
  • Turned early prototypes into repeatable internal systems by defining prompts, tools, state flows, validation steps, and operational handoff patterns.

Recognition

Third Prize
Qwen-Image LoRA Training Competition
12k+ Images
Curated ad dataset
40 Categories
Commercial product domains
7.47 Score
AI for Production track

Qwen-image-2512-lora-advertisement

Won Third Prize for a 1.7GB LoRA adapter that transforms standard product photos into commercial-grade ad visuals. The model was trained on 12k+ curated images across 40 advertising categories on a single consumer GPU.

What this shows: dataset curation, LoRA fine-tuning, production taste, GPU discipline, and rapid iteration under competition constraints.

Qwen Image LoRA competition AI for Production track

Qwen Image LoRA third prize leaderboard showing Rayyan Ahmed
Public recognition and proof
Tongyi Lab announcing Rayyan Ahmed as third prize winner ModelScope announcing Qwen Image LoRA competition winners

What I Ship

Reliable Agents

Multi-step workflows with explicit tools, state, retry paths, and validation layers instead of loose chatbot behavior.

Production AI

Systems that connect to real data, real portals, real users, and measurable outcomes.

Fast Iteration

From research prototype to usable workflow with clean evaluation loops and practical deployment constraints.

Technical Stack

Core technology stack

LangGraph MCP RAG LLM Ops Automation Qwen VLM

Featured Work

Solace AI

Advanced mental-health assistant focused on empathetic conversation, privacy-first architecture, and analytics for wellness workflows.

Stack: Transformers, therapeutic NLP, AWS, Redis, PostgreSQL, Docker, Kubernetes

View Solace AI

Agentic Document Extraction

Enterprise extraction system for messy scans, adaptive schemas, multi-record pages, cross-page duplicates, and resumable batch processing.

Stack: Qwen3-VL, LangGraph, Mem0, OpenCV, FastAPI, Celery

View Agentic Document Extraction

AutoApply AI

Full-stack job automation platform for discovery, resume scoring, tailored documents, human review, queues, WebSockets, and browser-based submission.

Stack: FastAPI, React, Playwright, Browser Use, Redis, LiteLLM, FAISS

View AutoApply AI

LinkedIn MCP Server

MCP server that gives LLM assistants structured LinkedIn job tooling for search, tailored resumes, cover letters, and application workflows.

Stack: Python, MCP, LinkedIn API, LLMs, FastAPI

View LinkedIn MCP Server

MLOps Pipeline - USD Volatility Forecasting

Production ML system that fetches live forex data, engineers 33 features, retrains every 2 hours, serves predictions through FastAPI, and monitors drift on each request.

Stack: Airflow, Grafana, Prometheus, MLflow, Docker, DVC, FastAPI

View MLOps pipeline

Medical Compliance Agent

Production healthcare AI workflow that reads clinical notes, extracts billing codes, and validates outputs through a LangGraph agent architecture with anti-hallucination checks.

Stack: Phi-3 fine-tuning, LangGraph, RAG, validation agents, healthcare automation

Production system at Nobility RCM

Explore all repositories

GitHub Analytics

GitHub streak Most used languages

Contribution graph

Daily Inspiration

Dev Quote

Connect

Available for work Open to collaborate Mentorship available

Production AI, practical automation, and measurable delivery.

Pinned Loading

  1. Solace-AI Solace-AI Public

    "Solace AI: Your Empathetic Digital Confidant": Solace AI is an empathetic mental health companion that understands your emotions and personality to provide personalized support through natural con…

    Python 17 6

  2. linkedin_mcp linkedin_mcp Public

    A powerful Model Context Protocol server for LinkedIn interactions that enables AI assistants to search for jobs, generate resumes and cover letters, and manage job applications programmatically.

    Python 18 5

  3. MLOps-Pipeline-for-USD-Forecasting MLOps-Pipeline-for-USD-Forecasting Public

    A production-grade MLOps pipeline for real-time USD volatility forecasting using EUR/USD forex data. This project demonstrates end-to-end ML lifecycle management with automated data ingestion, mode…

    Python 7

  4. AutoApply-AI-Agentic-Browser-Automation-for-Job-Search AutoApply-AI-Agentic-Browser-Automation-for-Job-Search Public

    AutoApply AI is an intelligent system designed to streamline your job search. It automates finding relevant job postings, crafting personalized resumes and cover letters using AI, and assists with …

    Python 24 7

  5. Agentic-Document-Extraction-PDF Agentic-Document-Extraction-PDF Public

    An agentic, zero‑shot document intelligence engine that sees, understands, and extracts from any PDF, no training, no hallucinations. Just define your fields and get trusted, structured outputs wit…

    Python 2

  6. ocr-app ocr-app Public

    State-of-the-art Optical Character Recognition (OCR) with Vision Language Model (VLM) integration for enhanced accuracy and optimal document processing.

    TypeScript 7 1