Nihal Vellaramkallingal Sulaiman
Data analyst and BI developer · Niagara Falls, Ontario
I build end-to-end data work in Python and SQL: pipelines and star-schema warehouses, data-quality checks, dashboards that managers open every morning, and forecasts and schedules that save money. I hold a Master's in Data Analytics from the University of Niagara Falls Canada and an MSc in Physics (astrophysics, with a mathematics minor) from Cochin University of Science and Technology.
Physics taught me to model a messy signal and check every number; analytics taught me to turn that into a decision someone can act on. I also know hotel operations from the floor, from over a year in hotel housekeeping in Niagara Falls. I am open to data analyst, BI developer and data science roles anywhere in Ontario.
news
- Nov 2026
- Started a blog with write-ups of my projects and Kaggle work, and moved this site to nihalsulaiman.com.
- Nov 2026
- Added an AI workflow kit to the Hospitality BI portfolio: safety rules for Claude Code on a live warehouse, an ETL triage method and a tested read-only SQL runner.
- Nov 2026
- Currently in the top 10% of 4,346 on the public leaderboard of Kaggle's Name That LLM competition.
- Oct 2026
- Graduated with a Master's in Data Analytics.
- Oct 2026
- Launched the Hospitality BI portfolio: a revenue command centre, a staff schedule builder and 12 hotel tools on a five-hotel sample group.
- Sep 2026
- Built WARL, a drone-swarm wildfire response with a live dashboard, at the GDG on Campus Build with AI hackathon.
selected projects
A synthetic five-hotel group with about 640,000 bookings over four years, loaded into a star-schema warehouse with ten automated data-quality checks. The Revenue Command Centre shows occupancy, ADR, RevPAR, booking pace against last year, a back-tested demand forecast and demand lost on sold-out nights. The staff schedule builder turns forecast workload into shifts with integer programming and checks them against Ontario labour rules.
One engine in which many small agents each make their own decision and together act as a coordinated team, tested against today's rules on hotel equipment, grid batteries and wildfire drones.
A decentralised drone swarm trained with a PyTorch DQN to route to fire choke points, with a live map dashboard and a Gemini chatbot that explains the swarm's decisions.
Machine learning, deep learning and optimisation competitions, including LLM identification, mass-spectrometry structure identification and smart-building forecasting. I keep honest hold-out sets and report where a result came from.
writing
2,391 of 2,000 free minutes, gone in about five hours. The billing detail I missed, and the guardrails I now put on every workflow.
My offline score said 0.930. The leaderboard dropped from 0.415 to 0.371. How I built a 300-molecule hold-out that finally agreed with the leaderboard.
Why I'm writing in public
Nov 1, 2026Who I am, what this blog will cover and why every post comes with the numbers and the parts that didn't work.
education
Master's, Data Analytics
2026University of Niagara Falls Canada, Ontario
MSc Physics
2017 – 2019Cochin University of Science and Technology, Kochi, India
Astrophysics major, mathematics minor.
BSc Physics
2013 – 2016Government Victoria College, Palakkad (University of Calicut), India
More: all projects · contact · site map · blog · now · free tools · CV · how I work