Vivek Narayana

Built a scalable customer-facing analytics pipeline using Temporal and Redis, cutting ingestion costs by 90%. Improved React dashboard query latency by 3–5x through optimized Apache Iceberg tables queried via Trino.

Collaborated with a startup to develop an intelligent assistant and safety app for the outdoors using Flutter. Developing modular RAG pipelines and multi-agent AI architectures implemented with AWS Bedrock. Optimized backend cloud infrastructure by cutting DynamoDB read latency by 90% with Redis caching.

Spearheaded the migration of Go-based microservices to AWS Lambda using Terraform and S3. Enhanced customer-facing front-end UI with React and integrated MongoDB to improve claim data accessibility. Collaborated on optimizing cloud resource usage to improve application performance and reduce costs.

Developed Python scripts to automate company link verification, improving workflow efficiency by 90%. Led the creation of RAG health chatbots with Streamlit and integrated AI tools like PineconeDB and Snowflake for vector search. Enhanced data visualization in Sisense and automated PDF-to-HTML conversion using Docker on EC2/S3, reducing manual effort by 4 months and ensuring system reliability through unit testing.

Taught programming to 100+ elementary students through Scratch, using custom game-based projects to reinforce essential coding skills. Achieved a 95% improvement rate in students' coding abilities by focusing on fundamental concepts like loops, variables, and functions.