Jeet Kishor Katariya

Developed a BERT-based semantic search for NLP queries into SQL/NoSQL queries with 93% match accuracy; deployed backend services on AWS EKS with Kubernetes; implemented OAuth 2.0/OIDC sign-in and JWT-based RBAC; reworked the data layer moving SQL workloads to DynamoDB; streamlined financial workflows with Celery processing ~43K records per run; automated TypeScript client generation from OpenAPI/Swagger for 250+ endpoints; managed CI/CD pipelines with Jenkins and Bitbucket for secure production rollouts.

Trained an LLM-based classifier to improve data relevance by 40% and tracked model performance with MLflow; designed a RAG chatbot with tuned chunking and prompts for multi-turn academic Q&A; enabled WebSocket with Socket.io for real-time feedback; configured FastAPI on EC2 with Nginx and Docker Compose; built an ETL pipeline to scrape 35 websites and integrate data into PostgreSQL; containerized workloads with Docker and implemented Grafana monitoring.

Developed an AI voice assistant with ElevenLabs and Twilio to automate front-desk calls with 86% intent accuracy; created a React & Spring Boot app to detect underperforming APIs and optimize backend performance by 57%; refactored high-traffic endpoint to offload concurrent writes via AWS SQS; integrated Kafka and CloudWatch with OTEL to expose ~5,000 events; built MapReduce pipelines on Hadoop to reduce reporting runtime by 35%; led upgrades for 52 SQL Server instances to reduce security exposure and licensing costs.