Headline

I build AI pipelines that cut manual compliance review time from weeks to hours across enterprise-scale codebases | Cisco SWE Trainee

About

Most compliance reviews on large codebases take engineers weeks, sometimes months, of manual grinding. Mine takes hours. I built a multi-phase AI pipeline — LangChain, ChromaDB, RAG, FastAPI — that scans large enterprise codebases for security compliance, replacing a manual audit process with an automated, evidence-backed workflow. Extensive automated test coverage, every output traceable back to source. My rule: AI can assist, but the output has to be deterministic and auditable — not a black box a security team has to trust blindly. I apply the same discipline to an automated testing framework I've built using Playwright and TypeScript. Stack: Python, LangChain, FastAPI, ChromaDB, RAG pipelines, Docker, TypeScript. Looking for backend or LLM engineering roles where I can own systems end-to-end. If you're building something similar, I'd like to compare notes.

Location

Madurai

AI stack

cursor

opencode

claudecode

Work experience

Software Engineer Trainee at Cisco, Oct 2025 - Present · 11 mos, Bengaluru, Karnataka, India, 1.Built and shipped a 12-phase agentic AI compliance platform (RAG + LangChain + JIRA integration) processing 168K+ files — cutting manual security review time from years to under 2 hours. 2.Designed and maintain a Playwright/TypeScript automated testing framework for Cisco's CUSP platform, improving test reliability and coverage. 3.Own code review standards across the automation initiative, enforcing consistency and maintainability at scale Working with Git branching strategies and CI workflows across a multi-engineer team.

Software Developer Intern at ideaForge, Aug 2025 - Sep 2025 · 2 mos, Navi Mumbai, Maharashtra, India, 1.Designed and developed a web automation system to sync product data from Excel to a company website, ensuring live, real-time updates whenever data changed. 2.Implemented two-way data reflection—updates made on the site were automatically logged back into Excel. 3.Worked with Full Stack Web Development Technologies(React,Node,Express,PostgreSql), to streamline data management workflows. 4.Collaborated with the R&D team to improve backend data processing efficiency and UI responsiveness. 5.Project successfully completed and stored on GitHub (deployment restricted to internal servers).

Java Development Intern at OctaNet Services Pvt Ltd., Jul 2024 - Aug 2024 · 2 mos, India, Developed and executed 3 innovative Java applications aimed at automating repetitive tasks, working approximately 15 hours a week.

Education

Bachelor of Technology - BTech, Information Technology, psna college of engineering and technology

12 th Grade, Computer Science, Mahatma Montessori Higher Secondary school

Public links

github: https://github.com/santhoshkumars2004