Rithvik Reddy Poreddy

Ai Engineer @Chevron

TX, US

Architected a production-grade LLM workflow with LangGraph and Python, orchestrating state-driven transitions that automated energy data processing and reduced manual handling.• Engineered a multi-step asynchronous ingestion pipeline that procesed multi-format inputs (.msg files, PDFs, and images) using OCR and text normalization delivering reliable downstream extraction and cutting processing delays.• Developed high-precision LLM extraction systems on Azure with Pydantic enabiling structured data extraction across multiple energy markets and improving data consistency for downstream analysis.• Designed a modular, scalable pipeline architecture using the Chain of Responsibility pattern, enabling maintainable integration across Python modules.• Built a specialized LLM-powered Excel parsing engine to convert complex, unstructured spreadsheets (.xlsx.xls) into standardized pricing and deal sheet models.• Implemented a rule-based validation and guidance layer in python to enforce business logic, perform utility and broker lookups, and automate approval workflows, resulting in faster approvals and fewer manual errors.• Integrated Azure functions Cosmos DB and Blob Storage to create a serverless architecture that improves stability and reduces operational overhead.• Established a test-driven development (TDD) framework with Pytest, adding extensive unit and integration tests and mock-service architecture, which increased code reliability and lowered regression bugs.• Deployed containerized services using Docker and GitHub Actions, automated CI/CD pipelines and delivering reliable production updates, which cut deployment time.

University of North Texas

Master of Science

Jawaharlal Nehru Technological University

Bachelor's degree