Devendra Babu Tati

Automated incident triage with an agent that parses Datadog logs and Java stack traces, cutting resolution time by about 70%. Rebuilt a multi-tenant FastAPI query service, reducing latency by 45% and cost per query by 30%.

Shipped the company's first AI product to production in 3 months, an end-to-end conversational assistant built on GPT-3.5, LangChain agents, and the Google Search API. Improved reply accuracy by 25% and response time by 40%.

Reduced LLM token overhead by 30% on a RAG chatbot for lease analysis and sped releases up 40% by automating CI/CD with Azure DevOps, Terraform, Docker, and Kubernetes. Boosted search accuracy and retrieval speed by 50% with a Python pipeline indexing lease documents into Elasticsearch.

Ran literature review, experiments, and write-up for two conference papers on memory allocation and reinforcement learning in large model training.

Shortened release cycles by 40% and reduced release errors with CI/CD pipelines in Jenkins, GitHub, and Maven. Standardised infrastructure across 3+ AWS accounts with reusable Terraform modules and remote state in S3 and DynamoDB.

Salesforce Agentblazer Champion certification achieved in 2025.

AWS Certified AI Practitioner certification, valid from August 2025 to August 2028.

Generative AI with Large Language Models certification completed in 2024.

Grade: 3.85/4.00

Master of Science in Computer Science with a concentration in Computational Data Science, achieving a GPA of 3.85/4.00.

Machine Learning certification from Stanford University.

Python for Data Science and AI certification completed in 2020.

Google IT Support Professional Certificate completed in 2020.

Databases and SQL for Data Science certification completed in 2019.