Muhammad Azam

AI Automation Engineer | Tech Lead | R&D

Driven to become a global leader in Autonomous AI systems. I specialize in building scalable intelligent agents, optimizing workflow automation, and bridging the gap between theoretical AI models and production-grade applications.

Technical Stack

AI Agents

Multi-Agent LLM

n8n

LangGraph

Next.js 16

TypeScript

Python

Claude API

OpenAI API

PostgreSQL

pgvector

Supabase

FFmpeg

Pine Script

DevSecOps

Automation

React 19

Prisma

Stripe

Docker

AWS (EC2, ECR)

GitHub Actions CI/CD

Vercel

Data Structures & Algorithms

Data Engineering

Experience

Summer 2026

Full-Stack & Cloud Engineering Fellow

Dev Weekends Fellowship 2026

Completed a structured curriculum across Engineering L0 & L1, Data Structures and Algorithms, Data Engineering, and Cloud & DevOps.
Applied it in two capstone projects (LMS Platform and Portfolio Site, listed under Completed Projects).

Sep 2025 - Jan 2026

Lead Automation Engineer

Confidential E-commerce Client

Engineered a full-stack autonomous agent capable of generating product titles, descriptions, and creative assets.
Optimized listing workflows for global marketplaces (Amazon, TikTok Shop), reducing manual entry time by 90%.
Deployed scalable n8n workflows to handle high-volume product data processing.

Aug 2025 - Sep 2025

Developer & Architect

Nexus AI Solutions

Built a high-precision recruitment intelligence tool for automated talent sourcing.
Implemented custom scraping algorithms to extract talent data from global repositories.
Integrated LLM-based qualification logic to score candidates against complex job descriptions.

Jun 2025 - Aug 2025

AI Automation Engineer

ReasonRail

Architected autonomous AI agent ecosystems to orchestrate complex business operations.
Workflow Strategy: Built state-based logic using LangGraph and n8n for autonomous decision-making.
Tactical Integration: Synced Gmail, QuickBooks, and Airtable to unify fragmented data streams.

Completed Projects

CoolCore — Heat Exposure Decision-Support

Built in 7 days for the FortyGuard Hackathon '26 (Team Coolcore). Turns FortyGuard's live heat-index and wet-bulb data into constraint-aware guidance for Louisiana residents who can't just "stay indoors" — a shift-exposure timeline, a shade-scored transit route comparator using streetview segmentation, and a supervisor-facing outdoor-worker safety board.

Next.js 16 TypeScript FortyGuard API MapLibre GL

View Live Demo

LMS Platform — Course Marketplace on Vercel & AWS EC2

Dev Weekends Fellowship project. Full-stack learning platform with Stripe Checkout and idempotent webhooks, Mux signed video playback, and Clerk role-based access (student / instructor / admin). One codebase ships two ways: Vercel, and a self-hosted Docker container on EC2 via a GitHub Actions → ECR → EC2 pipeline. Payment logic is covered by Vitest.

Next.js 16 Prisma Docker AWS EC2 GitHub Actions

View on GitHub

Portfolio Site with Case Studies

Dev Weekends Fellowship project. Dependency-free static site (plain HTML, CSS and JavaScript) deployed on Vercel: a home page, three deep-dive case studies (QuranicLens, FreshTrack, LMS Platform) and a downloadable resume.

HTML / CSS / JS Vercel Case Studies

View Live Site

Multi-Agent Knowledge Engine

Production-scale research platform orchestrating a 4-agent LLM pipeline (DeepSeek V3 → Claude Sonnet 4.6 → Claude Opus 4.6 + Gemini 2.5) over a 100,000+ record pgvector knowledge base. Streaming SSE responses, OAuth2 identity, strict grounding against verified embedded data. Built for Hackathon 2026.

Next.js 16 pgvector Multi-Agent LLM Claude Opus

View Live Demo

AutoDirector — Autonomous Video Pipeline

No-code n8n workflow that converts a single text prompt into a fully narrated, subtitled 1080p MP4 in under 10 minutes. Dual-LLM script generation (DeepSeek + Gemini 3), parallel TTS and generative-video workers (ElevenLabs + Fal.ai Kling), and FFmpeg assembly with fallback-first error resilience.

n8n ElevenLabs Fal.ai FFmpeg

FreshTrack — Enterprise Inventory PWA

Mobile-first multi-tenant PWA for restaurant operations featuring batch-level FEFO tracking, barcode scanning (EAN/UPC/QR), waste analytics, and moving-average reorder forecasting. Architected on Next.js 16 Server Actions with Drizzle ORM and role-based Supabase Auth.

Next.js 16 Drizzle ORM PostgreSQL PWA

Digital Sentinel — Zero-Trust Agent Infrastructure

Hardened deployment architecture for autonomous reconnaissance agents operating in hostile network environments. Rootless Podman containment, Tailscale mesh perimeter, UFW firewall, and dual-tier AI routing (Groq for rapid triage, Claude for deep reasoning) under Telegram-based human-in-the-loop C2.

Podman Tailscale DevSecOps Zero Trust

SMC Trading Ecosystem

Production Smart Money Concepts indicator in Pine Script v5 (~770 LOC) featuring BOS/CHoCH structure tracking, Order Blocks, Fibonacci OTE zones, and Quasimodo reversal detection — no repainting. Shipped alongside a 7-agent Claude Code skill system that accelerates Pine Script R&D end-to-end.

Pine Script TradingView Claude Skills

Bachelor of Science in Computer Science

Virtual University • Focus on Software Engineering & AI

Final Year Project: Smart Query Routing & Email Automation System (University Intelligent Communication Hub)

DeepLearning.AI · Stanford University

Coursera

📜

Statistics with Python Specialization

University of Michigan

Coursera

📜

Programming for Everybody (Python)

University of Michigan

Coursera