Ahmed Usman
Ahmed Usman

AI Systems Engineer

AhmedUsman

AI/ML EngineerAI Product Development, LLM Apps & RAG

Building AI products that combine LLMs, LangGraph, FastAPI, retrieval, automation, and strong user-facing product execution.

PakistanOpen to AI/ML, LLM, and backend engineering opportunities

Main Stack

LangGraphLangGraph
ClaudeClaude
FastAPIFastAPI
OpenAIOpenAI
PineconePinecone
WhisperWhisper
TwilioTwilio
AWSAWS
Ahmed Usman

Focus

AI product development across LLM workflows, retrieval systems, agent orchestration, backend APIs, and user-facing product delivery.

FocusAI Products
StackLangGraph + FastAPI

About Me

Shipping AIthat people can use

Ahmed Usman

Ahmed Usman

AI/ML Engineer

ahmadusman050@gmail.com
+92 335 0707006
Pakistan

Profile

AI/ML Engineer focused on AI product development across production-grade LLM applications, RAG systems, agentic workflows, voice automation, and applied computer vision. I work across backend architecture, model orchestration, integrations, and product-facing frontend delivery to ship AI software that is reliable, scalable, and actually useful to end users.

AI product developmentProduction LLM systemsRAG and agent orchestrationVoice agents and automation

AI Product Engineering

From model workflows to frontend delivery, I build AI products that solve real user problems instead of stopping at prototypes.

Workflow Automation

I connect models with business systems, APIs, schedulers, and human review loops so automation becomes reliable and operationally useful.

RAG & Agent Systems

I design retrieval, routing, caching, and orchestration layers that make AI apps more grounded, observable, and scalable.

MY JOURNEY

Professional Experience

Associate AI Engineer

Tech Emulsion
Jul 2025 - Present
Pakistan
  • Design and deploy LLM-powered conversational agents, RAG pipelines, and automation workflows for production AI products.
  • Build backend services and integrations using Python, Django, FastAPI, n8n, AWS, and third-party APIs.
  • Collaborate with product, engineering, and operations teams to deliver scalable AI systems from concept to deployment.

AI/ML Engineer

DevK System
Aug 2024 - Jun 2025
Pakistan
  • Built LLM applications including RAG systems, autonomous agents, and agentic workflow prototypes.
  • Designed and deployed machine learning systems with attention to latency, scalability, and maintainability.
  • Applied MLOps practices including Dockerized deployment, CI/CD workflows, and model monitoring.

Back-End Developer

Brandora
Sep 2022 - Jul 2024
Pakistan
  • Developed REST APIs using Node.js, Express.js, and MongoDB for production web backends.
  • Integrated backend services with React frontends to support reliable application data flow.
  • Implemented authentication, reusable modules, and query optimization across application features.

Core Stack

Tools behind thesystems I ship

LangGraph

LangGraph

LangChain

LangChain

OpenAI

OpenAI

Anthropic

Anthropic

Pinecone

Pinecone

ChromaDB

ChromaDB

Whisper

Whisper

pgvector

pgvector

Selected Projects

Production workacross AI and automation

AVL Copilot

Completed

A production multimodal RAG and technical-support copilot for the audio, video, and lighting industry, built to help field technicians troubleshoot equipment, retrieve manuals, and diagnose issues from text and images.

Built a strictly sequential LangGraph pipeline with semantic cache, intent detection, RAG, routing, web search, and SSE streaming.
Implemented dynamic token budgets, per-user quotas, live observability endpoints, and multimodal image diagnostics.
FastAPILangGraphPineconeRedisSupabaseStripeOpenAI

The Meatery AI Platform

Completed

An internal operations and AI voice-agent platform for a US premium-meat e-commerce brand, combining analytics, CRM automation, inventory logic, and self-improving Retell AI call workflows.

Engineered inbound and outbound voice-agent workflows with live cart lookup, discount generation, SMS delivery, and call safety guardrails.
Built nightly transcript-analysis automations that used Claude to generate rebuttals, battle cards, and prompt-improvement recommendations.
Next.jsTypeScriptPrismaPostgreSQLRetell AIn8nShopify

Good Food Project

Completed

A full-stack AI content engine for a UK organic food brand with human-in-the-loop review, brand-grounded content generation, image matching, and scheduled publishing.

Created an async job-based generation pipeline with provider routing, token-cost controls, and RAG over brand knowledge.
Delivered reviewer workflows, quote-image composition, image editing, and automated publishing through Buffer.
FastAPISupabaseLangGraphNext.js 15React 19Cloudflare R2

LinkedIn Outreach Automation Platform

Completed

A vision-driven automation platform that runs a real-time LinkedIn outreach pipeline from scraping and enrichment through AI note generation and connection sending.

Built a three-tier AI decision cascade to handle LinkedIn UI variants using action-bar snapshots, screenshots, and deterministic label scanning.
Implemented multi-tenant sessions, retries, cancellation, resume flows, and connection analytics.
FastAPIReact 19SupabasePlaywrightQwenDashScope

Lost-N-Find

Completed

A mobile AI lost-and-found platform with claim scoring, admin review, realtime notifications, and semantic matching for noisy or incomplete item descriptions.

Combined deterministic matching with LLM semantic scoring to improve claim verification quality.
Added real-time updates, JWT auth, item matching, and graceful AI fallback behavior.
FastAPISupabaseReact NativeExpoWebSocketsMistral

WhatsApp Freight Agent

Completed

A multimodal AI freight-procurement workflow that automates RFQ intake, vendor outreach, quote collection, follow-up, and booking confirmation from WhatsApp conversations.

Handled text, voice, PDF, and cargo-image inputs in one operational workflow.
Reduced token cost by replacing redundant LLM calls with keyword routing, code-based extraction, and templates.
n8nWhatsApp APIOpenAIWhisperGmail APIGoogle Sheets

ACADEMIC JOURNEY

Educational Background

B.S. Electrical Computing & Communication Engineering

University of Engineering and Technology, Peshawar
Peshawar, Pakistan
Completed

Certifications

Continuedlearning

Machine Learning Specialization

Coursera

Completed

Deep Learning Specialization

Coursera

Completed

Machine Learning in Production

Coursera

Completed

TensorFlow for AI, ML and DL

Coursera

Completed

Generative AI with LangChain and Hugging Face

Udemy

Completed

Contact

Let's buildsomething useful

Availability

Open to AI/ML, LLM, and backend engineering opportunities. If you're hiring for AI engineering, LLM products, RAG systems, voice automation, or backend-heavy product work, I'd be happy to talk.

Ahmed Usman

Open to AI product, LLM engineering, and backend-focused collaboration.

© 2026 Ahmed Usman. All rights reserved.