Kumar Suman
Lead Data Scientist
Bengaluru, Karnataka, India
Summary
👤
Senior🎓
Top SchoolKumar Suman is a Senior Data Scientist based in Bengaluru with nine years of experience building high-impact, revenue-driving AI systems across media, retail, and supply chain. He has led end-to-end personalization and feature-store efforts for 15M+ users at Dainik Bhaskar and architected low-latency pipelines processing 100M+ daily events. At Noodle.ai he created Inferno, a forecasting platform that outperformed traditional projections and saved clients millions, and earlier built distributed feature engineering frameworks that increased deployment velocity and compute efficiency by orders of magnitude. Most recently he designed a vector search engine with LLM-powered RAG capabilities at Neuron7.ai and now applies that production ML expertise at Victoria’s Secret & Co. Comfortable bridging research and engineering, Kumar pairs deep ML/DL skills in Python with pragmatic system design to take prototypes into production at scale. An interesting thread through his work is squeezing extreme performance—latency, throughput, and cost—out of real-world ML systems rather than just improving model metrics.
9 years of coding experience
7 years of employment as a software developer
Bachelor’s Degree, Information Science and Engineering, Bachelor’s Degree, Information Science and Engineering at PES UniversityPython Specialization , Computer Science, Python Specialization , Computer Science at University of MichiganMachine Learning & Deep Learning Specialization, Computer Science, Machine Learning & Deep Learning Specialization, Computer Science at Coursera
Github Skills (65)
python10
flask-application10
coreference-resolution10
deep-learning10
flask10
modular10
ml10
neural-network10
algorithm10
stock-price-prediction10
artificial-intelligence10
electron10
typescript10
tensorflow10
natural-language-processing10
Programming languages (7)
Github contributions (5)
nityansuman/warren
Apr 2019 - Nov 2022
Web app to predict closing stock prices in real time using Facebook's Prophet time series algorithm with a multi-variate, single-step time series forecasting strategy.
Contributions:6 releases, 4 reviews, 108 commits in 3 years 8 months
time-seriestime-series-forecastingstock-price-predictionneural-networksfbprophet
dunovank/dark-jupyter
Oct 2018 - Nov 2018
Dark Theme for Jupyter Notebook
Contributions:48 commits in 1 month
jupyter-notebook
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.