I'm Hakkeem, a Full Stack AI Engineer.
I work as an R&D Engineer (Intern) at a Fin-Tech startup, building AI agents for fraud detection and deploying production systems like an AI copilot for real-time banking data queries. 🤖
My expertise spans production ML deployment (CPU optimization, binary embeddings for RAG latency), payment gateway architecture with Kafka/PostgreSQL, and taking POCs from research to production. 💻🔧
Currently exploring graph techniques for enterprise fraud detection, optimizing inference latencies, and learning MLOps pipeline automation. ☁️
Email | LinkedIn | Twitter | Stack Overflow
- “Applying Monolithic to Microservices Strategy for Elastic Container Deployment of AI Applications” presented at the 2025 IEEE 6th Annual World AI IoT Congress (AIIoT), Seattle, USA.
- “MLOps Challenges in Deploying High-Performance Vision Models: An Empirical Analysis” presented at the IEEE CONECCT 2025, Bangalore.
- huggingface/smolagents#1462
- Infisical/infisical#1406
- Infisical/infisical#1407
- ultralytics/yolov5#12785
| Project Name | Description | Technologies | Link |
|---|---|---|---|
| RAG-Powered Case Similarity and Complaint Search | Developed a RAG-powered complaint search system that retrieves similar complaints based on semantic similarity and geographical location, incorporating modus operandi analysis to detect related cases based on victim patterns, crime locations, and suspect behavior. Integrated CCTV camera mapping to find nearby surveillance footage for rapid law enforcement investigation. Presented to Kochi City Police to demonstrate how Generative AI and LLMs can assist police officers in case analysis and crime prevention. | RAG, OpenAI, Embeddings, Semantic Search, Geospatial Analysis, ChromaDB, FastAPI | GitHub Repo |
| AccentAce | Accent Ace is an AI-powered software that generates text for users and records their voice to analyze their pronunciation. The goal of this project is to enable users to improve their pronunciation skills independently. | Next.js, FastAPI, Gemini, Firebase, Docker, GitHub Actions | Live Demo |
| Distributed Object Detection on Edge Devices using tinyML | Developed an object detection system that identifies cars from camera streams and compares them to a target image, facilitating vehicle tracking across distributed devices with edge-based processing. [Academic] | Python, Yolo, Object Detection, Raspberry PI, MQTT, tinyML | GitHub Repo |
| College CMS | A Content Management System for colleges built with Laravel, PHP, MySQL, and JavaScript. It features an admin dashboard for content management and multiple user roles controlled access. | php, Laravel, MySQL, JavaScript | GitHub Repo |
- Moving from Request-Response to Async: Engineering Scalable Notification for Better Performance
- From EC2 to RDS: Mastering Database Migration with AWS DMS
- The Make-a-Ton Journey: Building Kerala’s Largest Student Hackathon
- Map-in-Review: An Analysis of your Travel from Google History