Ahammad Nafiz
Associate Product Engineer, DeepDive at Markopolo AI · Dhaka, Bangladesh · --:--
- I'm an AI engineer in Dhaka, working as an Associate Product Engineer at Markopolo AI. I work on DeepDive, an AI social listening platform, plus an AEO visibility tool and an AI data analysis agent.
- Before that I owned the eval service at AskTuring.ai and built the retrieval and benchmarks for a production RAG system.
- I learn by rebuilding things: an autograd engine in NumPy, a memory layer for agents, a research agent that cites its sources. Lately I keep drifting toward mechanistic interpretability.
- Next: a PhD in Europe, then a lab working on the frontier.
01Work
6 projects
Atlas
Source control for coding agentsContributorCore contributor to Atlas, an open-source agentic IDE (Rust, Tauri, React, 9.4k+ stars) that links every git commit to the AI coding agent session that wrote it.
- Rust
- Tauri
- React
- HNSW
Buyziz
AI commerce platformLiveA multi-tenant e-commerce SaaS with an AI sales agent that sells in Bangla and English. I built the agent and the store builder, and led product engineering.
- LLM agents
- Tool calling
- Multimodal
- Multi-tenant SaaS
fizzlabai
Deep research agentIn developmentA deep research agent built with LangGraph and Gemini. It runs a fixed 8-phase pipeline over the open web and returns a cited Markdown answer with a full trace.
- Python
- FastAPI
- LangGraph
- Gemini
Engram
Persistent memory for AI agentsOpen sourceLong-term memory for AI agents on PostgreSQL and pgvector. 90.8% on LongMemEval-S, 94.1% on LoCoMo-10 and 79.7% on BEAM 1M, with no LLM calls at ingest.
- Python
- PostgreSQL
- pgvector
- BM25
Personal Knowledge Assistant
Multi-agent RAG systemOpen sourceA multi-agent RAG system for your own books and notes. Low-confidence answers fall back to web search automatically.
- Python
- FastAPI
- FAISS
- Gemini
FizTorch
Deep learning framework from scratchOpen sourceA PyTorch-style deep learning framework built from scratch in NumPy: an autograd engine, core layers and an Adam optimizer.
- Python
- NumPy
- Autograd
02Experience
Markopolo AINow
Aug 2026 – PresentAssociate Product Engineer, DeepDive
Working on DeepDive, an AI social listening platform, plus an AEO (answer engine optimization) visibility tool and an AI data analysis agent, across a TypeScript and Python LLM stack.
AskTuring.ai
Sep 2025 – May 2026Jr. Software Engineer
Owned the LLM evaluation service and admin dashboard, automating RAG benchmarks on HotpotQA, MuSiQue, FinQA and CUAD. Built Deepen Answer, a production RAG pipeline: extended-thinking LLMs, an agent that fills its own knowledge gaps, and hybrid BM25 plus vector retrieval with reciprocal rank fusion. Built FastAPI services that improved retrieval accuracy and cut query latency.
United International University
Dec 2025 – Feb 2026Research Assistant
Researched autonomous AI agents and multi-agent systems, and contributed to papers on LLM applications, retrieval and agentic workflow design.
03Recognition
- Oct 2025
Champion, IC6 University Innovation Hub
BDT 60,000 pre-seed after a 15-week program
- Jan 2025
Best Prototype, Blockchain Olympiad
Team Auctra, UIU CSE Fest. BDT 20,000 prize
- Jun 2024
Champion, CSE Project Show
Predicta, an end-to-end data analysis and ML platform
- 2024
1st Runner-Up, Data Innovators Challenge
Sales analytics track
04Contact
Hiring an AI or machine learning engineer, want to talk research, or have a good problem? My inbox is open.
ahammadnafiz86@gmail.com