Associate AI/ML Engineer · Mayfair Asian Food Industries

Islam Nabi

AI/ML Engineer | Agentic AI Developer

Building production-grade AI systems that ship.

Download Resume Contact Me

3+ Years of Experience in AI

5+ Years of Experience in Software Engineering

7+ AI Projects

5 AI Domains

Live Production Deployments

01 — Get To Know More

About Me

I'm Islam Nabi, an Associate AI/ML Engineer passionate about intelligent systems that solve real-world business problems. My work spans Agentic AI, LLMs, RAG, Speech AI, and Computer Vision, with a focus on shipping solutions from concept to production. I design and deploy AI applications combining reasoning, tool use, autonomous decision-making, and scalable cloud infrastructure building multi-agent systems, speech pipelines, and computer vision products that deliver measurable business impact. I hold a BS in Computer Science from Sukkur IBA University and explore multi-agent architectures, advanced RAG, LLM fine-tuning, and enterprise AI automation.

My philosophy is simple:

"AI doesn't replace humans, It amplifies what we can do."

Location Lahore, Pakistan

Years of Experience 2+ Years in AI/ML Engineering

Current Role Associate AI/ML Engineer @ Mayfair

Education BS CS — Sukkur IBA University

Specialization

Agentic AI LLMs RAG Speech AI Computer Vision

Availability Open to collaborations on Agentic AI systems, LLM applications, and production ML pipelines.

02 — Professional Journey

Experience Timeline

Nov 2025 – Present

Associate AI/ML Engineer

Mayfair Asian Food Industries Ltd. · Lahore

Dec 2025 – Mar 2026

AI Trainer – Code & Reasoning Evaluator

Revelo · Remote

Sep – Nov 2025

AI/ML Engineer Intern

PureLogics Software House · Lahore

May – Jul 2025

AI/ML Engineer Trainee

PureLogics Software House · Lahore

Jul – Aug 2024

Software Engineer Intern

InternnCraft · Remote

Nov 2023 – Apr 2024

LLM Trainer

Remotasks (San Francisco, USA) · Remote

03 — Featured Work

Featured Projects

Production-grade AI systems designed for real-world impact and scalability.

AI Sales Intelligence System

ETL pipeline ingesting 150,000+ daily records into PostgreSQL across 12 materialized views

ReAct Text-to-SQL LangChain LangGraph FastAPI Next.js PostgreSQL

ETL pipeline ingesting 150,000+ daily records into PostgreSQL across 12 materialized views
Real-time KPI dashboard with sales trends, regional breakdowns, and top SKU performance
Autonomous Text-to-SQL ReAct agent — fully grounded queries with zero hallucinations
Hardened SQL layer: SELECT-only enforcement, row limits, and retry caps

View on GitHub

AI-Based HS Code & Landed Cost Estimation

Production RAG over FBR tariff database — ~95% coverage across 7,500+ indexed HS codes

RAG FastAPI AWS FAISS LangChain OpenAI FBR Customs Tariff

Production RAG over FBR tariff database — ~95% coverage across 7,500+ indexed HS codes
8-step FBR-compliant classification engine with ~90% accuracy
Compound customs duty calculator with full AED→PKR landed cost breakdown
Deployed on AWS EC2 (FastAPI + Gunicorn/Uvicorn + Nginx)

View on GitHub

Urdu Text-to-Speech System

100-hour clean Urdu speech dataset with automated data prep pipeline

Whisper VITS XTTS-v2 LoRA ECAPA-TDNN Silero VAD FastAPI PyTorch

100-hour clean Urdu speech dataset with automated data prep pipeline
GPU-accelerated Urdu transcription via Whisper Large v3, benchmarked with WER
Fine-tuned VITS, XTTS-v2, Chatterbox, and Tortoise TTS using LoRA
Zero-shot voice cloning via ECAPA-TDNN speaker embeddings

View on GitHub

Employee Attendance System (CCTV-Based)

Real-time face recognition on live CCTV via InsightFace — >95% accuracy, multi-camera

InsightFace Buffalo Model FAISS PostgreSQL Next.js OpenCV

Real-time face recognition on live CCTV via InsightFace — >95% accuracy, multi-camera
FAISS-powered identity matching with cosine similarity for instant attendance
Automated attendance logging to PostgreSQL on every recognition match
Next.js HR dashboard with one-click Excel attendance export

View on GitHub

Smart Vehicle Insurance Risk Assessment

End-to-end MLOps: ingestion, validation, training, S3 model registry, live inference

Scikit-learn FastAPI MongoDB Atlas AWS S3/EC2/ECR Docker GitHub Actions

End-to-end MLOps: ingestion, validation, training, S3 model registry, live inference
XGBoost classifier for claim likelihood prediction, versioned in S3
Full CI/CD via Docker, GitHub Actions, AWS ECR — zero-downtime EC2 deploys

View on GitHub

Wrist Abnormality Detection System

Trained YOLOv9 on GRAZPEDWRI-DX (20,327 X-rays, 9 pathology classes) — ~70% mAP

YOLOv9 LLMs Flask Medical Imaging GRAZPEDWRI-DX

Trained YOLOv9 on GRAZPEDWRI-DX (20,327 X-rays, 9 pathology classes) — ~70% mAP
LLM-powered diagnostic reporting turning detections into clinical summaries
Flask app: X-ray upload, real-time inference, bounding boxes, PDF download

View on GitHub

Visitor Management System

Enterprise system with role-based access (Admin/Receptionist) and session auth

Next.js 16 React 19 TypeScript MongoDB AWS S3 Tailwind CSS 4

Enterprise system with role-based access (Admin/Receptionist) and session auth
Real-time operations dashboard with live visitor/vehicle stats
Cloud-native biometric pipeline: webcam capture → S3 → pre-signed URLs
CNIC blacklist enforcement, QR-coded badges, and CSV exports

View on GitHub

04 — Frontier Focus

Technical Expertise

Production capabilities across agentic workflows, speech AI, computer vision, and MLOps — the differentiating layer of my work.

Agentic Workflows

LangChain, LangGraph, ReAct Agents, Tool-Use Agents

RAG Pipelines

FAISS, Vector DBs, Hybrid Search, Contextual Retrieval

LLM Orchestration

LangChain, Prompt Engineering, Function Calling, Chains

Speech AI Systems

Whisper (ASR), VITS, XTTS, VAD, Speaker Embeddings

LLM Fine-Tuning

LoRA, PEFT, Hugging Face Transformers, WER Evaluation

AI Deployment

FastAPI, Streamlit, Flask, Docker, AWS EC2/S3

MLOps & Monitoring

MLflow, CI/CD Pipelines, GPU Inference Optimization

Computer Vision

YOLOv9, OpenCV, CNNs, Transfer Learning, Face Recognition

05 — Core Stack

Tech Stack

Languages, frameworks, and infrastructure I use to build and ship AI systems.

Languages

Python C++ JavaScript TypeScript Java SQL

AI / ML / DL

PyTorch TensorFlow Keras Hugging Face Transformers Scikit-learn XGBoost Pandas NumPy CNNs RNNs LSTMs Transformers Transfer Learning Fine-Tuning LoRA PEFT LLM Fine-Tuning NLP Computer Vision Object Detection (YOLO) Feature Engineering Model Evaluation Classification Regression Clustering

Generative AI & LLMs

Prompt Engineering Retrieval-Augmented Generation (RAG) LLM Fine-Tuning Embeddings Vector Databases (FAISS, ChromaDB, Weaviate, Milvus) Function Calling Structured Outputs Multi-Modal AI AI Assistants

Agentic AI

AI Agents Multi-Agent Systems Agentic RAG ReAct Agents LangGraph LangChain Workflow Orchestration Tool Calling Planning & Reasoning Agents Autonomous Decision-Making Systems Human-in-the-Loop Workflows C-Level AI Agent Systems Conversational AI Agents

MLOps & Deployment

Docker Git GitHub MLflow DVC FastAPI Flask REST APIs AWS (EC2, S3) CI/CD Model Deployment Model Monitoring

Databases & Data Engineering

PostgreSQL MongoDB SQL ETL Pipelines Data Warehousing Polars Pandas Data Processing Data Validation

06 — Get In Touch

Let's Build Something Together

Open to collaborations on Agentic AI systems, LLM applications, and production ML pipelines.

Send a Message