Hello, I’m
Backend Software Engineer · Node.js & Machine Learning
Backend engineer with 3+ years of commercial experience building Node.js APIs, payment systems and high-volume data pipelines for fintech, government and hospitality clients. Currently completing an MSc in Computer Science at the University of West London, with first-author deep learning research presented at the FARSY 2026 Faringdon Symposium.
I am a backend software engineer based in Isleworth, London, with over three years of commercial experience at Ekbana Solutions & Tuna Technology. I have designed and built Node.js APIs, payment systems and high-volume data pipelines for fintech, government and hospitality clients across the United States, Japan and Nepal — including an invoice automation microservice that cut manual data entry by 80% and a booking backend handling more than 10,000 reservations a month.
Alongside industry work, I am completing an MSc in Computer Science at the University of West London, where I serve as Course Representative. My dissertation research applies deep learning to InSAR-based volcano monitoring: I am first author of a paper presented at FARSY 2026, the 1st Faringdon Symposium, in which a CNN-LSTM model exceeded a published research benchmark.
I also bring four years of teaching experience — as a part-time lecturer in .NET and as a teaching assistant and module tutor on Birmingham City University partnership programmes — mentoring dozens of final-year students each intake on software architecture, API design and database modelling. I care about measurable delivery, clear communication and building systems that hold up under real-world load.
Sep 2025 – Sep 2026 (expected)
University of West London, UK
Graduated Jan 2022
Sunway International Business School (IUKL, Malaysia)
Jan 2022 – Aug 2025
Ekbana Solutions & Tuna Technology
Jun 2024 – Aug 2025
G.P. Koirala Memorial College, Kathmandu
Jan 2021 – Oct 2024
Sunway International Business School, Kathmandu
Professional — Fintech
Node.js microservice parsing 1,000+ PDF invoices per day with automated NetSuite synchronisation for a US fintech client — reducing manual data entry by 80%.
Professional — Payments
Mono-repo payment hub integrating multiple providers (eSewa, ConnectIPS, Khalti) with end-to-end encrypted transaction APIs, secure webhook handling and Redis/BullMQ queues for asynchronous processing.
Professional — Hospitality
Booking backend for a Japanese hospitality platform serving 5 hotels and 10,000+ monthly reservations, with lottery and priority-sale allocation logic, third-party availability API integration and PostgreSQL queries optimised for peak load.
Professional — Government
Content-management REST APIs and notification queues for the Nepal Electricity Authority’s citizen-facing mobile app — a national government portal.
Research — MSc Dissertation
ResNet-50 and CNN-LSTM classifiers trained in PyTorch on the Thalia benchmark (6,131 multispectral satellite time-series samples, 44 volcanoes), improving F1 by 6.2 percentage points over the baseline and exceeding the published benchmark.
Solo Project — 2020–2021
Chrome extension backed by a Flask ML API. Trained five classification models on ~20,000 news articles using TF-IDF features, selecting LinearSVC at 96% accuracy, with cross-referencing against trusted outlets via the Google Custom Search API.
Bibek Gautam (first author) — presented at FARSY 2026, the 1st Faringdon Symposium (Sensing the Impact), London, June 2026.
MSc dissertation research at the University of West London (Jan 2026 – present), co-authored with researchers from the THRIVE Centre and the Faringdon Research Centre for Non-Destructive Testing and Remote Sensing. Implemented and trained ResNet-50 and CNN-LSTM classifiers in PyTorch on the Thalia benchmark — 6,131 multispectral satellite time-series samples across 27 channels and 44 volcanoes — and designed a shuffled-frame ablation study isolating per-frame encoding, rather than chronological ordering, as the source of the performance gain.
I am open to backend engineering opportunities, research collaboration and interesting projects. The quickest way to reach me is by email — I aim to reply promptly.