AI Engineer. Data Scientist. Builder.
Based in Bengaluru
I build AI that understands enterprise data, from the model to the product people use.
01The engineer behind the work
Model. Data. Product.
01The engineer behind the work
Model. Data. Product.
I build AI systems that sit on top of real enterprise data and answer the questions people actually ask.
Since 2019 I have shipped websites for clients, earned an MSc in Big Data Analytics, built NL2SQL and Graph RAG pipelines at NuMinds AI, and now build enterprise AI applications at Eurofins.
Give a question a path to the answer.
LLM applications that reason over enterprise data: NL2SQL that turns plain language into correct queries, Graph RAG that retrieves context with structure, and chatbots people trust.
Find the signal before building on it.
Forecasting, recommendation, and analytics grounded in statistics: ARIMA sales forecasts, cold-start recommenders, and big data pipelines on Spark and Databricks.
Put the model where people work.
Five years of client websites taught me that AI only matters once it ships: Python backends, Flask and Streamlit apps, Next.js frontends, dashboards, and cloud deployment.
02Journey
Since 2019.
2019
Started building websites for clients: UI, SEO, and backends, while still in school.
2020
BCA at Kristu Jayanti College, Bengaluru, alongside freelance web work.
2023
MSc in Big Data Analytics at St. Joseph's University: statistics, ML, Spark, and forecasting.
2024
Built carbon footprint reporting at Deepwoods Green Initiatives.
2025
NL2SQL, domain chatbots, and Graph RAG pipelines at NuMinds AI.
Today
Building AI-powered enterprise applications at Eurofins IT Delivery Center.
Next
Systems that let anyone ask their data a question and trust the answer.
03Selected work
Questions, answered.
03Selected work
Questions, answered.
Three projects that show the range: language models on enterprise data, NLP for public good, and machine learning that works from the first visit.
LLM systems NuMinds AI, 2025
Fine-tuned NL2SQL models and Graph RAG pipelines that let non-technical teams question enterprise data directly, with domain chatbots for HR analytics and customer insights built on top.
LLM + NLP Personal project
A conversational app for Bengaluru that maps a reported crime to the relevant IPC laws in real time, giving people actionable guidance instead of legal jargon.
Machine learning Personal project
A movie recommender that solves the cold-start problem with content-based filtering, so new users get relevant picks before they have any history.
Time series / Personal project
ARIMA forecasting of future sales to help optimize inventory and reduce stockouts.
Data product / Deepwoods Green Initiatives
Scope 1, 2, and 3 emissions reporting with dashboards and third-party API integrations for carbon accounting.
Agentic AI / Personal project
A multi-agent writing pipeline built with crewAI.
Computer vision / Personal project
A CNN weather classifier served through a Streamlit app.
Web development / Freelance, 2019 to 2024
Responsive sites and dashboards for clients, including Urooj, Street Cars, and a healthcare management app.
04How I work
Problem first.
04How I work
Problem first.
The same discipline in every project, tuned to what the problem needs.
01
Start from the decision a user needs to make and the data that can support it.
02
Connect the LLM to schemas, graphs, and retrieval so answers are traceable, not invented.
03
Measure accuracy on real queries, then deploy behind APIs the business already uses.
01
Profile sources, clean what is broken, and name what the data cannot tell us.
02
Pick the simplest model that captures it, from time series to CNNs.
03
Report results in the units the business uses: accuracy, precision, stockouts avoided.
01
Design the interface around the user's task, not around the model.
02
Backends, APIs, and responsive frontends that are fast and maintainable.
03
Ship to the cloud, watch real usage, and iterate.
05Contact
Hiring for AI engineering, or have data that should answer questions? I read every message.