# Muhammed Rasin O M

> AI and data engineer with 1.5 years of industry experience building LLM backends and data pipelines in Python: remote backend work for DearPriceHunter, a Swedish startup, and data engineering at JMAN Group. Hands-on with document ingestion, RAG, LLM evaluation and REST APIs. Founder of Misata, an open-source synthetic data tool with 50+ users and one pilot running.

Open to full-time AI and data roles in engineering, product or go-to-market. Remote, on site or relocating. Roles: AI Engineer, Machine Learning Engineer, Data Engineer. Also open to: Forward Deployed Engineer, Solutions Engineer, GTM Engineer, Technical Product Manager. Based in Kozhikode, Kerala, India (Asia/Kolkata). Working in industry since July 2024.

Contact: rasinbinabdulla@gmail.com. GitHub: https://github.com/rasinmuhammed. LinkedIn: https://www.linkedin.com/in/rasinmuhammed. X: https://x.com/RasinMuhammedX. Medium: https://medium.com/@rasinbinabdulla. PyPI: https://pypi.org/user/rasinmuhammed/. ORCID: https://orcid.org/0009-0009-1971-7291.

## Profile
- [Profile, experience and projects](https://www.rasinmuhammed.tech/): the full profile on one page
- [CV](https://www.rasinmuhammed.tech/cv): printable, also as PDF
- [Why hire me](https://www.rasinmuhammed.tech/hire): the story so far, from school to now, and the team I am looking for
- [CV as JSON Resume](https://www.rasinmuhammed.tech/resume.json)

## Experience
- Software developer, contract, DearPriceHunter AB (Dec 2025 – Aug 2026): A Swedish startup that analyses what companies spend with their suppliers. I worked remotely on the backend and its LLM features.
- Backend and AI developer, Independent (Mar 2025 – Present): Freelance work for Onepane AI, and my own open-source projects, mainly Misata.
- Software engineer, data engineering, JMAN Group (Jul 2024 – Feb 2025): My first job after graduating, in the data engineering team of a data and analytics consultancy. The synthetic data work there is where Misata started.

## Projects
- [Misata](https://github.com/rasinmuhammed/misata): Synthetic data. Generates realistic test data with several linked tables from a short description of a business. If you say revenue should grow from 10,000 to 50,000 over a year, the generated rows add up to exactly that, and every link between tables is valid. Meets all 136 checks in its public test suite, where a typical Faker script meets 78. Misata Studio, the hosted version, has more than 50 users and one pilot, and was selected for Claude for Startups.
- [Ratio](https://github.com/rasinmuhammed/ratio): Legal search. Answers legal questions about Indian court judgments, with a citation for every claim. It searches 10,588 judgments and tells apart what the court decided from what a lawyer argued. Finds the exact case asked for 89% of the time on 313 questions I labelled by hand, up from 59% with keyword search and 4% with semantic search alone.
- [Entify](https://github.com/rasinmuhammed/entify): Duplicate detection. Finds duplicate customer records in messy data, such as the same person entered twice with a typo. It sets itself up from the data, explains why it matched two records, and needs no specialist tuning. On the FEBRL benchmark, every pair it flagged was a real duplicate, and it found 98.7% of them.
- [Nimbal](https://github.com/rasinmuhammed/nimbal): Energy forecasting. Forecasts how much power wind and solar farms will produce the next day, as a range of likely outcomes rather than a single number, and turns the forecast into bids for the UK electricity market. Replayed against the 2024 HEFTCom forecasting competition, it would have placed 2nd of 66 teams on accuracy and 1st on trading revenue.
- [Harara](https://github.com/rasinmuhammed/harara): Heat safety. Hour-by-hour work and rest plans for outdoor workers in the Gulf, based on heat forecasts and Qatar’s heat-stress rules. Compared on the same hours, the ECMWF forecast missed 12 to 18% of the hours when work should have stopped; NOAA GFS missed 40 to 45%.
- [Mitate](https://github.com/rasinmuhammed/mitate): Recommendations. Product recommendations for small online shops with a few hundred products and a few dozen regular customers. It works from the catalogue on day one and improves as orders come in. Beat the best single method at every shop size on the public OTTO shopping dataset.
- [Argus](https://github.com/rasinmuhammed/industrial-copilot): Maintenance assistant. An assistant for factory maintenance teams that shows how close a machine is to failing. Every number it gives is calculated by code, not written by the language model. Passes all of its test cases on the UCI AI4I 2020 machine-failure dataset.
- [Footnote](https://github.com/rasinmuhammed/footnote): Source checking. Checks AI search answers against the web pages they cite. Every claim must point to a source, and every source is actually read. Runs can be replayed from saved results without new searches.
- [Yaqzan](https://github.com/rasinmuhammed/yaqzan): Disaster response. A disaster-response simulation based on Kerala’s 2018 floods. An AI model sends boats and trucks where they are needed, and a checker blocks any order that names a resource that does not exist. Built for the MBZUAI and IFM K2 showcase.
- [The Quantum Canvas](https://rasinmuhammed.github.io/quantum-generative-art/): Generative art. Type a name and it becomes a small planetary system, drawn from the results of a five-qubit quantum circuit. Runs in the browser with Qiskit and p5.js.

## Open source
- [misata](https://pypi.org/project/misata/): The Python library behind Misata. Describe the totals a dataset must hit and get rows that add up to them exactly, with a report you can rerun to check.
- [indicfuzz](https://github.com/rasinmuhammed/indicfuzz): Matches Indian names that sound alike but are written differently, across scripts, word order, initials and titles, for KYC checks and record matching. Scores 0.987 ROC AUC on its test set, against 0.699 for the popular rapidfuzz library.
- [rag-tui](https://pypi.org/project/rag-tui/): A terminal tool for checking how documents are split up for retrieval (RAG), with live previews, batch tests and export to LangChain or LlamaIndex.
- [opsdeck](https://pypi.org/project/opsdeck/): An admin panel for FastAPI apps that use async SQLAlchemy, with permissions, per-row access rules and audit logs.
- [pytest-misata](https://pypi.org/project/pytest-misata/): A pytest plugin that gives tests consistent, linked multi-table data from Misata in one line.
- [mangograd](https://pypi.org/project/mangograd/): A readable autograd engine in NumPy, tested against PyTorch.
- [node-canon](https://github.com/rasinmuhammed/node-canon): Finds and merges duplicate entries in knowledge graphs built by language models, with no API key needed.

## Writing
- [The Museum of Finished Things](https://www.rasinmuhammed.tech/writing/the-museum-of-finished-things): On reverse thinking. Start from the thing as it finally stands, trace it back to whatever brought it into being, and the world turns into a museum.
- [All writing](https://www.rasinmuhammed.tech/writing)

## Optional
- [Declarative Outcome-Conformant Synthesis: Exact, Closed-Form Specification Satisfaction and a Conformance Benchmark](https://arxiv.org/abs/2606.08736): arXiv:2606.08736, the method behind Misata
- [RSS](https://www.rasinmuhammed.tech/rss.xml)