Resume

EXL Services · India

Own the deployment and development pipelines for computer-vision and generative-AI systems serving insurance risk and claims processing — from GPU inference optimisation through to production serving on AWS.

• Built and optimised deployment and development pipelines for PyTorch aerial-imagery models, focusing on GPU-accelerated inference, memory-efficient tensor operations and scalable infrastructure to support high-throughput image analysis for insurance risk and claims processing.
• Designed and implemented end-to-end Python pipelines covering data ingestion, embedding generation, RAG pipelines, vector search and inference for enterprise applications.
• Used ONNX to optimise deep-learning model exports, cutting deployment overhead and reducing the runtime memory footprint.
• Deployed predictive models to production through both batch pipelines on Airflow and real-time pipelines on AWS Fargate, coordinating delivery across the data science and IT teams.
• Designed and built the data pipeline that derives flags from claim notes, which serves as the central feature repository for downstream predictive models.
• Implemented a serverless architecture on API Gateway, AWS Lambda and DynamoDB, with deployment artefacts served from S3.
• Developed Spark SQL scripts in Python to accelerate large-scale data processing.
• Partnered directly with data science teams at insurance clients to build AWS data-pipeline solutions, and handle day-to-day operational issues and performance tuning of live applications.

Tiger Analytics · India

Built the PySpark analytical ecosystem that unified first-party and third-party data into model-ready datasets for downstream analytics.

• Designed and developed an analytical ecosystem framework pipeline in PySpark that combines first-party and third-party data sources and delivers model-ready data and insights.
• Designed and developed end-to-end ETL solutions and processing applications using Spark and Hive.
• Developed the Python and PySpark jobs that load data into the analytical ecosystem table schema on a monthly interval.
• Built Tableau dashboards for the quality-check process and automated that process for seamless flow.
• Contributed to an agile development team focused on data ingestion across multiple sources.

Mindtree · India

Built internal tooling, data-processing CLIs and Flask microservices, and shipped the first production ML model of my career.

• Designed and developed a Python tool that automatically captures product test failures and opens a bug for each one, backed by an auto-generated regex for the failure signature.
• Modelled a classifier in Python to predict the competitor for a given segment, and deployed it as a Flask API on AWS.
• Developed more than 10 Flask API microservices for backend systems.
• Developed command-line tools in Spark and Python that let customers process data faster and communicate insights.
• Optimised query performance through index forcing, constraint-based loading and related techniques.
• Performed front-line code reviews for other development teams, and supported UAT testing with reporting for business users.