Pavun Kumar R
AI Engineer architecting autonomous workflows & secure LLM systems
Building systems,
not wrappers
I am an AI Engineer specializing in building production-grade agentic workflows and secure LLM architectures. I do not build simple wrappers — I architect autonomous systems that solve complex backend friction.
Currently at ThoughtSeed, I focus on bridging the gap between raw LLM capabilities and reliable business applications. My core stack includes Next.js, LangGraph, Supabase, and AWS.
Technical Highlights
- ▶Architecting stateful, human-in-the-loop autonomous agents using LangGraph
- ▶Built and deployed a Generative AI model for processing and analyzing historical legal judgments, optimizing data pipelines with PostgreSQL
- ▶Developing secure LLM architectures tailored for high-compliance environments
I am actively building in public and documenting my code. Always open to discussing technical architecture, remote engineering opportunities, or micro-SaaS development.
Where I’ve shipped code
ThoughtSeed
1 yr 6 mos · Hybrid
Associate Innovation Developer
Full-time · Jul 2025 – Present
Junior Innovation Developer
Full-time · Mar 2025 – Jul 2025
Full Stack Developer
Internship · Dec 2024 – Mar 2025
LUNEWISE Technologies
Director · On-site
Led software infrastructure and engineering leadership at a technology startup in Madurai, Tamil Nadu.
WhatsLegal.ai
AI & Data Engineer · Remote (Germany)
Built AI/ML pipelines and data engineering solutions for a legal-tech platform based in Germany.
Tech stack & tools
AI & LLM
Backend
Frontend
Cloud & Infra
Data & Tools
What I’ve built
EmployAI — Open Source AI Agents Platform
Open SourceSelf-hosted AI agents that auto-reply to customer emails and generate LinkedIn posts. Gmail auto-reply with knowledge base context, LinkedIn content generation with approval workflow, multi-tenant architecture, and swappable LLM backends (Groq, LM Studio, OpenAI).
Autonomous Legal AI Agent
Generative AI model for processing and analyzing historical legal judgments. Optimized data pipelines with PostgreSQL for high-throughput document ingestion.
Secure LLM Architecture
Production-grade secure LLM architecture tailored for high-compliance environments. Stateful, human-in-the-loop autonomous agents with full audit trails.
RAG Pipeline System
Retrieval-Augmented Generation pipeline bridging raw LLM capabilities with reliable business applications. Production-deployed with automated monitoring.
Let’s connect
Open to discussing technical architecture, remote engineering opportunities, or micro-SaaS development. Reach out — I respond fast.