Open to opportunities

Muhammad Ahmad Fraz

@muhammadahmadfraz

Message

AI Engineer and LLM Architect building production-grade GenAI, RAG, and computer vision systems that ship reliably.

Pakistan

Message

What I'm looking for

I’m looking to lead GenAI and AI platform projects—turning research-grade work into reliable, cost-efficient products—where strong MLOps, evaluation (RAGAS/TruLens), and cross-functional teamwork are valued.

I’m an AI Engineer and Full-Stack developer with 8+ years of experience designing and shipping production-grade AI systems—from LLM-powered agents and RAG pipelines to real-time computer vision and predictive analytics. I’m currently a Tech Lead at AB Ark, architecting GenAI solutions for clients across fintech, healthcare, logistics, and e-commerce.

I focus on the end-to-end ML lifecycle: data engineering, model fine-tuning, multi-modal inference, cloud deployment, and MLOps at scale. I’m known for translating research-grade AI into reliable, cost-efficient products, including multi-tenant AWS SaaS analytics, LLM agent platforms using LangChain/LangGraph, and RAG-based clinical knowledge experiences.

Across roles, I’ve delivered measurable outcomes—cutting manual processing by 80% with document intelligence pipelines, achieving <1.2s response latency with streaming RAG, and deploying YOLOv8 systems with 92% mAP on-device. I also established CI/CD and evaluation frameworks (MLflow, GitHub Actions, RAGAS, TruLens), and mentored a cross-functional team of 6 engineers to improve deployment speed and quality.

Experience

Work history, roles, and key accomplishments

Current

Tech Lead - AI Engineering

Current

AB Ark Private Limited

Feb 2023 - Present (3 years 4 months)

Architected a multi-tenant AWS analytics SaaS using Django/FastAPI/PostgreSQL for 10+ enterprise clients with 99.9% uptime. Built production LLM agent and RAG platforms, including a RAG mental-health app with <1.2s latency and a document intelligence pipeline that cut manual processing time by 80%, and established MLOps to reduce deployment from 2 weeks to 1 day.

AWS EC2 Lambda S3 RDS Django fastAPI RAG (Pinecone)MLOps GitHub Actions Docker Kubernetes Tensorflow Lite YOLOv8

Senior AI Engineer

Tkbees

Mar 2022 - Oct 2022 (7 months)

Built an AI resume parser using spaCy NER with BERT fine-tuning, reaching 94% entity-extraction accuracy, and integrated it with a React/Node.js hiring portal. Implemented a synthetic avatar pipeline for automated call-handling using DeepFace and Coqui TTS, reducing call-center costs by 35%, and deployed ML microservices on Azure AKS, AWS ECS, and GCP Cloud Run with Terraform IaC.

React DeepFace Coqui TTS DeepSpeed Azure AWS ECS

Data Scientist

Al Badr Engineering

Mar 2020 - Feb 2022 (1 year 11 months)

Trained and deployed YOLOv5 + Faster-RCNN ensemble for real-time manufacturing defect detection, reducing false-negative rate to under 2%. Developed a Django quality-control portal and LSTM/GRU time-series forecasting (MAPE <6%), and performed competitor pricing scraping/analysis with Scrapy and Pandas to generate actionable insights.