Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I'm Bhargob Deka, a senior AI/ML developer focused on forecasting, anomaly detection, and AI agents who loves turning research into scalable production systems. I collaborate with cross-functional teams to deliver cloud-native solutions that empower decision-making and operational excellence. I also write technical articles and case studies to help startups adopt practical AI, sharing learnings from projects at Hydro-Québec and Jesta I.S.…Hi, I'm Bhargob Deka, a senior AI/ML developer focused on forecasting, anomaly detection, and AI agents who loves turning research into scalable production systems. I collaborate with cross-functional teams to deliver cloud-native solutions that empower decision-making and operational excellence. I also write technical articles and case studies to help startups adopt practical AI, sharing learnings from projects at Hydro-Québec and Jesta I.S.

Bhargob Deka

Data Scientist, AI Developer, Full Stack Developer, +2





Hi, I'm Bhargob Deka, a senior AI/ML developer focused on forecasting, anomaly detection, and AI agents who loves turning research into scalable production systems. I collaborate with cross-functional teams to deliver cloud-native solutions that empower decision-making and operational excellence. I also write technical articles and case studies to help startups adopt practical AI, sharing learnings from projects at Hydro-Québec and Jesta I.S.…Hi, I'm Bhargob Deka, a senior AI/ML developer focused on forecasting, anomaly detection, and AI agents who loves turning research into scalable production systems. I collaborate with cross-functional teams to deliver cloud-native solutions that empower decision-making and operational excellence. I also write technical articles and case studies to help startups adopt practical AI, sharing learnings from projects at Hydro-Québec and Jesta I.S.

Available to hire

Hi, I’m Bhargob Deka, a senior AI/ML developer focused on forecasting, anomaly detection, and AI agents who loves turning research into scalable production systems. I collaborate with cross-functional teams to deliver cloud-native solutions that empower decision-making and operational excellence.

I also write technical articles and case studies to help startups adopt practical AI, sharing learnings from projects at Hydro-Québec and Jesta I.S.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

Senior AI-ML Developer at Jesta I.S.

October 1, 2024 - Present

Led the design and development of the company's first AI Agent enabling retail clients to obtain business insights from natural language queries against their own data. Prototyped with Gradio, built a scalable backend with FastAPI and a frontend using React and TypeScript. Deployed via Azure DevOps CI/CD, implemented I/O guardrails, and used Langsmith for experiment tracking and evaluation. Engineered end-to-end demand forecasting with LightGBM models for short-term and pre-season forecasts, orchestrated with Kedro, runs on AWS Batch, and tracked experiments with MLflow. Developed AutoEncoder and Random Forest models for a fraud-detection POC.

Technical Writer at Unstract & Literally.dev

January 1, 2025 - Present

Crafted engaging blog posts and case studies for two AI startups, promoting products like LLMWhisperer and Kapa’s production-grade RAG pipelines for technical documentation API and MCP server, helping drive audience growth and lead generation.

Machine Learning Researcher at Hydro-Québec

December 31, 2022 - April 30, 2024

Developed probabilistic LSTM models for forecasting using data from 400+ dams; integrated a real-time anomaly detection framework with switching Kalman filter, reducing false alarms by ~50%. Built an end-to-end PyTorch framework for training, evaluation, and monitoring, integrated with MLFlow on AWS; improved predictive accuracy by modeling heteroscedastic uncertainty.

Graduate Research Assistant at Polytechnique Montreal

September 1, 2018 - December 1, 2022

Developed a multiplicative state-space model to capture non-harmonic periodic patterns in structural health monitoring; introduced AGVI for real-time uncertainty quantification in Bayesian ML models, reducing training time.

Graduate Research Assistant at McGill University

September 1, 2015 - March 1, 2018

Enhanced statistical downscaling methods by integrating multiple techniques to project future climate scenarios; evaluated and optimized statistical models for daily/hourly downscaling.

Teaching Assistant at McGill University

August 31, 2016 - March 31, 2017

Led tutorial sessions for Probabilistic Systems and Engineering System Analysis for 120 undergraduate Civil & Mining Engineering students.