Open to full-time roles May 2026

Hi, I'm .
I translate numbers
into ML models
that actually work.

Data Scientist & ML Engineer with 4+ years transforming high-volume data into production-ready models, from time-series forecasting and causal inference to RAG pipelines and LLM systems built for regulated environments.

M.S. Data Science · University of Maryland · 3.5+ yrs industry experience at Infosys · Generative AI Intern at Centennial Technologies · Open to Data Science, AI/ML, and Risk roles

Get in Touch

About

The person behind
the pipelines.

I'm a Data Scientist with a rare combination of deep technical fluency and a genuine obsession with whether the model is actually trustworthy. My career started at Infosys, where I spent 3+ years building and validating production ML systems, not just shipping models, but stress-testing them, flagging exposures, and making sure the decisions they informed were sound.

I'm currently completing my M.S. in Data Science at the University of Maryland, where I've also done research in multimodal AI and built production RAG systems grounded in scientific telemetry. I'm drawn to roles where data science intersects with real business stakes, whether that's ecosystem analysis, model risk, or AI-powered product development.

Outside of work I'm passionate about making technical insights legible, the kind of communication that moves organizations, not just reports.

M.S. Data Science

University of Maryland, College Park

May 2026 GPA 3.7

B.E. Computer Science

Rashtrasant Tukadoji Maharaj Nagpur University

May 2020 GPA 3.8

Skills

What I work with.

Statistical & ML Modeling

Multivariate RegressionRegression & ClassificationXGBoostSARIMAXTime-Series ForecastingCausal InferenceA/B TestingCUPEDBayesian StatisticsSHAP / LIMEPropensity Score Matching

Programming Languages

PythonSQLRJavaScriptJavaC++

Data Engineering & Cloud

PySparkApache SparkSpark & HiveBigQueryDatabricksSnowflakePostgreSQLETL PipelinesAWS S3 / LambdaAWS BedrockAWS SageMakerGCP

Generative AI & LLMs

RAG PipelinesRAG SystemsLangChainLLaMA 2Azure OpenAILLMs & GenAINLP & TransformersPrompt EngineeringLLM EvaluationAgentic AICrewAIFAISSPyTorchTensorFlow

MLOps, Infrastructure & Development Tools

MLflowGitGit & CI/CDDockerREST APIs

Visualization, Analytics & Product Metrics

TableauPower BIStreamlitKPI / OKR Design

Experience

Where I've worked.

May – Aug 2025

Generative AI Intern

Centennial Technologies · Tysons, VA

Designed real-time RAG pipelines and Agentic AI architecture using Azure OpenAI with long-term memory and tool-use modules, built LLM-as-a-judge evaluation workflows establishing output quality benchmarks for mission-critical regulatory and medical device compliance systems.

May – Aug 2025

Research Assistant

MSP Lab, University of Maryland · College Park, MD

Productionized a custom RAG architecture over large-scale oceanographic datasets for the Sensorium AI project and automated multimodal inference pipelines using PyTorch and R, reducing latency for real-time voice-to-voice human-AI interaction.

Aug 2024 – Present

Teaching Assistant - Database Design

University of Maryland · College Park, MD

Mentored 350+ students across 100+ assignments in SQL optimization, ER modeling, and normalization.

Oct 2022 – Aug 2024

Data Scientist II

Infosys Limited · Hyderabad, India

Built end-to-end ETL pipelines on Databricks/Spark processing ~2TB daily, engineered 40+ predictive features using Spark SQL, developed SARIMAX and multivariate regression models improving forecast accuracy by 19%, executed 15+ A/B tests and causal inference experiments, integrated outputs via REST APIs into real-time KPI dashboards, and scaled distributed ML pipelines via MLflow driving $2M+ in planning decisions for 200+ stakeholders.

Mar 2021 – Oct 2022

Data Engineer

Infosys Limited · Hyderabad, India

Migrated legacy asset management data to a Hadoop/Spark big data lake consolidating 6+ source systems, built Java microservices handling 150K+ daily requests cutting infrastructure overhead by 22%, and designed data warehouse schemas with table partitioning, SQL query layers, and Tableau/Power BI reporting dashboards for 10+ internal consumers.

Projects

Things I've built.

RAG-Powered Document Intelligence System

End-to-end RAG pipeline over 10K+ documents using LangChain and FAISS. Deployed as a document Q&A system on Hugging Face Spaces with Streamlit UI, supporting semantic search across PDF corpora.

↓ 34% hallucination rate 10K+ documents

LLaMA 2LangChainFAISSRAGASStreamlit

LLaMA 2 Medical Chatbot

Healthcare AI chatbot with a vector database-backed RAG pipeline. Features a custom benchmarking framework to evaluate reasoning quality and response stability under high-load simulations.

↑ 22% answer relevance ↓ 28% latency

LLaMA 2LangChainRAGVector DBPython

Network Traffic Forecasting & KPI Platform

SARIMAX time-series forecasting models and a feature engineering pipeline (40+ features, Spark SQL) powering real-time KPI dashboards consumed by stakeholders.

↑ 19% forecast accuracy $2M+ business impact

SARIMAXSpark SQLDatabricksMLflowTableau

Music Playlist Recommender & Mood Classifier

Classifies songs into moods using Spotify audio features like valence, tempo, energy, and danceability, then recommends similar tracks via K-Means and HDBSCAN clustering with cosine similarity scoring.

5 mood categories K-Means + HDBSCAN

Pythonscikit-learnHDBSCANStreamlitSpotify API

Image Color Palette Extractor

Extracts dominant colors from any uploaded image using K-Means clustering and generates a Pinterest-style color palette with HEX/RGB values, color harmony suggestions, and WCAG accessibility contrast ratings.

3–10 color extraction WCAG AA/AAA support

PythonK-MeansStreamlitPillowscikit-learn

Hierarchical Cross-Modal BRCA Subtype Classifier

Multimodal deep learning framework fusing histopathology images and gene expression data to classify BRCA subtypes using a hierarchical cross-modal attention architecture.

Multimodal fusion Cross-modal attention

PyTorchMultimodal AIAttentionGenomicsPython

A/B Testing & Causal Inference Framework

End-to-end experimentation framework implementing A/B testing with statistical significance testing, propensity score matching, and causal inference techniques to isolate true treatment effects from observational data.

95%+ confidence intervals Causal inference

PythonCausal InferencePropensity ScoringSciPyPandas

Customer Transaction Forecasting

Time-series forecasting pipeline predicting customer transaction volumes using SARIMAX and XGBoost ensemble models, with automated feature engineering from transactional history and seasonal decomposition.

SARIMAX + XGBoost Time-series forecasting

PythonSARIMAXXGBoostPandasscikit-learn

Hi, I'm .
I translate numbers
into ML models
that actually work.

The person behind
the pipelines.

What I work with.

Where I've worked.

Things I've built.

Let's connect.

My creative space.

Thanks for scrolling
till the end.

Hi, I'm .I translate numbersinto ML modelsthat actually work.

The person behindthe pipelines.

What I work with.

Where I've worked.

Things I've built.

Let's connect.

My creative space.

Thanks for scrollingtill the end.

Hi, I'm .
I translate numbers
into ML models
that actually work.

The person behind
the pipelines.

Thanks for scrolling
till the end.