Aman Gupta

Associate Data Scientist

Bangalore, Karnataka 2+ yrs exp 85 · Excellent

About

Associate Data Scientist with 2+ years of hands-on experience in data analytics, machine learning, statistical analysis, and AI-driven solutions. Skilled in Python, SQL, Snowflake, exploratory data analysis, data visualization, model evaluation, and data quality validation. Experienced in translating complex business requirements into reliable analytical solutions, developing regression testing frameworks, and delivering actionable insights for business-critical applications. Additional experience in Generative AI, semantic modeling, MCP, AtScale, and AI-agent development, including building and validating AI-powered assistants.

Skills & Expertise (38)

Generative AI Advanced

8.5/10

Years Exp

Python Advanced

8.0/10

Years Exp

NLP Advanced

8.0/10

Years Exp

SQL Advanced

7.5/10

Years Exp

Git Advanced

7.0/10

Years Exp

AWS S3 Intermediate

6.5/10

Years Exp

Power BI Intermediate

6.0/10

Years Exp

Tableau Intermediate

6.0/10

Years Exp

Docker Intermediate

6.0/10

Years Exp

scikit-learn Matplotlib PyTorch TensorFlow GCP Azure Azure AI Foundry GitHub Actions Azure DevOps Jira Postman Confluence Fine-tuning C++ Postgresql MySql Snowflake MS Excel LLMs AI Agents Prompt Engineering FastAPI RAG Model Evaluation MCP Semantic Models Streamlit Plotly Flask

Work Experience

Associate Data Scientist

Data Radian Technologies

Jun 2024 - Present

AI-Powered Analytics Assistant & RAG Validation Framework: Built, tested, and deployed 2+ AI-powered analytics assistants using Snowflake Cortex, AtScale semantic models, and MCP to provide scalable natural-language access to analytics across 1B+ records. Built a RAG pipeline with Snowflake Vector Search for document-based Q&A and grounded LLM responses. Implemented chunking, embeddings, semantic retrieval, and context prompting to improve relevance and reduce unsupported answers. Designed golden datasets and regression test suites to evaluate AI assistant performance using metrics such as answer accuracy, expected-result match, consistency, and response time, validating structured analytical responses against source data before UAT and production releases. Integrated structured data retrieval through AtScale with contextual document retrieval through RAG so the assistants could handle both analytical questions and knowledge-based queries within a single workflow. Automated Data-Quality Workflows for BUS_VAULT: Developed automated data-quality workflows in Snowflake using SQL to validate data inserts and updates, detecting duplicate records, null values, and duplicate training data across the BUS_VAULT. Achieved 90% coverage of data-quality checks across the vault, improving data integrity and reliability while reducing manual validation effort. AI Evaluation Framework for Sentiment & Classification Models: Implemented TruLens-based AI evaluation workflows combining human evaluation and automated function-based checks for sentiment analysis and classification. Achieved 90% agreement between automated TruLens evaluations and human evaluation, with 85–90% accuracy on sentiment classification. Turbine Device Data Analytics: Owned the end-to-end data analytics workflow for 5K+ broken turbine devices, leveraging SQL to identify failure patterns and deliver actionable operational insights. Collaborated with business teams to improve issue identification and resolution, supporting data-driven decision-making.