Aman
Bhatt
Data Scientist & Machine Learning Engineer specialising in forecasting, real-time ML systems and MLOps. Proven ability to deliver production-grade models with <3% MAPE, architect low-latency data pipelines with Spark, Kafka and AWS, and design CI/CD-driven ML workflows for scale and reliability.
Data Scientist & ML Engineer
Data enthusiast and a proficient data scientist with a background in civil engineering. Currently working as an Assistant Manager at BSES Rajdhani Power Limited, applying advanced algorithms and techniques to solve complex forecasting problems.
Passionate about leveraging data-driven insights and building AI systems that last — driving measurable business impact across the energy sector. Experienced in bridging research with engineering, enabling grid operators and renewable energy firms to make data-driven decisions at speed.
Skills & Tools
From raw data to deployed model — the full ML engineering toolkit.
Work Experience
Building real ML systems in the energy sector.
- Leading the design and deployment of short-term and day-ahead load forecasting models, enhanced with an LLM-powered assistant built using LangChain that allows grid operators and planners to query demand forecasts and scenario analyses in natural language, improving decision-making speed, transparency, and operational efficiency across BRPL's service zones.
- Led the development of predictive models for Delhi Discoms (BRPL, BYPL, TPDDL), achieving a 3% MAPE using Neural Networks, LightGBM, CatBoost, and XGBoost for day-ahead and intraday forecasting. Integrated real-time weather data while leveraging Apache Spark and Kafka for efficient data handling.
- Developed solar and wind forecasting models for REMC-Telangana (166 plants), designing high-precision week-ahead energy predictions. Deployed models using AWS SageMaker, with automated CI/CD pipelines for seamless updates.
- Conducted extensive research and development on price forecasting at IEX, employing diverse models and achieving a notable 7.5% MAPE with innovative strategies using the LightGBM model.
- Designed foundation for different equipment and structures that are usually present in oil refineries using softwares like AutoCAD, StaadPro, RISA Base, Foundation 3D, RAM Connection etc.
My Work
End-to-end ML projects — all deployed as live apps.
A movie recommendation system that recommends movies based on search.
IPL win predictor that can predict probability of winning of chasing team at any point of the match.
Analysis of past years summer olympics record from 1896 to 2016.
A classifier that can predict any SMS as spam or not.
Analyse of group as well as personal WhatsApp chats.
Let's build something
that ships to production
Full-time roles · Contract work · Consulting · Research collaborations in ML, AI and data science.
Get in Touch
Have a role or project in mind? Fill in the form — I'll reply within 24 hours.
Let's connect
Open to full-time ML/AI roles, contract work, consulting and research collaborations.