Headline

Data Scientist @ ADA | 1x Databricks & 4x Snowflake Certified | Scalable Data Pipelines • ML Solutions • Cloud Analytics |Ex-Infoscion

About

With 4.2+ years of experience in Data Science and Data Engineering, I specialize in designing scalable data pipelines, building AI/ML models, and delivering analytics solutions across e-commerce, supply chain, financial services, and marketing domains. I am skilled in Databricks, Snowflake, PySpark, Kafka, Airflow, and AWS, with expertise in Python, SQL, TensorFlow, and scikit-learn. My work includes optimizing pipelines and dashboards to cut processing times by 60%+, automating API-to-Delta workflows, and developing ML models for forecasting, marketing mix optimization, and document intelligence. Certified in Databricks Data Engineer Associate and SnowPro Core, I bring both engineering depth and applied AI expertise to help organizations turn complex data into scalable systems and actionable insights.

Location

Bengaluru

AI stack

claude code

claude

gemini

hugging face

notebooklm

huggingface

Work experience

Data Scientist at ADA, Apr 2024 - Present · 2 yrs 6 mos, Bengaluru, • Developed a Agentic RAG-based chatbot on 5K+ unstructured documents in S3 using LLM , enabling contextual question answering and reducing manual data search time by 70%. • Built an agentic AI system using LangGraph that generates SQL queries from natural language questions and executes them on Databricks, reducing analyst query load by 60%. • Implemented a Power Automate API triggered from Databricks to generate CSV reports, enabling 100+ non-technical users to self-serve curated datasets without SQL knowledge, reducing report request time by 80%. • Built and maintained a data engineering and ML pipeline by first fetching WMS data via API into S3, then using PySpark streaming to ingest then build an ML model on the data and update Delta tables with optimization in Databricks for Tableau, Superset, and FastAPI+Dask backend API services for React dashboards, reducing manual handling by 80% and improving load times by 65%. • Developed a deep learning–based revenue forecast–driven profit and loss prediction model that enhanced financial accuracy by 25% and optimized strategic planning. • Created a Teams chatbot with Snowflake Cortex Analyst for natural language querying, reducing reliance on analysts and saving 10+ hours/week.

AI/ML Engineer at Kramasoft_MortgageSpecialist, Feb 2023 - Mar 2024 · 1 yr 2 mos, Bengaluru, Karnataka, India, • Built APIs to gather data from multiple financial institutions and store it in S3, creating a unified data source for downstream document classification and analytics. • Developed a Python-based backend application for document classification and trimming, using ML models to detect document types and extract relevant pages ; additionally, applied OCR on scanned documents using AWS Textract for text extraction. • Integrated the application with Google Gemma on AWS Bedrock to generate accurate answers from extracted content and predefined questions. • Created a Docker image and deployed the application on AWS Lambda, enabling scalable, serverless execution.

System Engineer - Data & AI at Infosys, Jun 2021 - Jan 2023 · 1 yr 8 mos, Bengaluru, Karnataka, India, • Implemented AWS-based data engineering workflows using S3, Glue, Lambda, and Athena, enabling scalable data ingestion, transformation, and storage to support analytics solutions. • Built a demand forecasting model on OMS data using neural networks in Databricks, improving sales and inventory planning accuracy. • Developed a product affinity model with PySpark ML FP-Growth, boosting cross-sell and upsell strategies by identifying frequently bought-together items. • Designed a buyer model linking demand, orders, and customer behavior, enabling more accurate customer growth forecasting. • Optimized marketing analytics workflows by building a Marketing Mix Model with TensorFlow and PyMC, enabling data-driven budget allocation and improving ROI analysis by 30%. • Optimized all data models in Databricks with PySpark + SQL pipelines by implementing incremental refresh (instead of full refresh), reducing pipeline refresh time by 40% and Power BI dashboard reload time by 30%, leading to significant time and compute cost savings.

Summer Intern at Defence Research and Development Organisation (DRDO), May 2019 - Jun 2019 · 2 mos, Balesore, Orissa, India, • Focused on understanding pre-built machine learning models used in missile flight safety as part of the GNSS project. • Studied a model designed to predict if a missile would crash, learning about the input features it utilized. • Analyzed another model responsible for determining the crash location, which outputs precise latitude and longitude coordinates. • Gained valuable insights into how these models function, including their inputs, outputs, and their role in missile trajectory analysis and safety protocols.

Education

Bachelor of Technology - BTech, Electronics and Communications Engineering, C.V.Raman College of Engineering

+2, Pcmit, College of basic science and humanities Bhubaneswar