I build intelligent agents
that think and act.
Agentic AI Engineer at Toyota Motor North America — building multi-agent orchestration, RAG pipelines, and production LLM systems at enterprise scale.
Featured Work
Projects
LangGraph · GPT-4o · Pinecone · FastAPI
Multi-Agent Research Assistant
Multi-agent system with specialized sub-agents (researcher, critic, synthesizer) that autonomously retrieves, validates, and synthesizes information. RAG pipeline with Pinecone hybrid dense-sparse search and re-ranking. Deployed on AWS Lambda with streaming output.
LangChain · ReAct · Tool Calling · PostgreSQL
Autonomous Customer Support Agent
Agentic customer service bot using LangChain ReAct framework with custom tools — order lookup, refund processing, CRM updates, Twilio notifications. Sliding-window conversation memory with entity extraction. Guardrails AI safety layer for PII detection.
Spring Boot · Kafka · LlamaIndex · Weaviate
RAG E-Commerce Assistant
Event-driven order processing with Spring Boot and Kafka, augmented by a RAG-powered AI assistant using LlamaIndex + Weaviate vector store for product recommendations and order status explanations. Kubernetes HPA autoscaling.
Flask · Redis · WebSocket · GPT-4
Global Chat Platform with AI Moderation
Real-time chat service with WebSocket support, Redis Pub/Sub messaging, and GPT-4 AI moderation layer for content classification and filtering. Multi-region failover with DynamoDB global tables.
What I Build
Expertise
From agent design to production deployment — I architect and build end-to-end AI systems for enterprise environments.
Multi-Agent Orchestration
Designing autonomous agent systems with LangGraph and LangChain that coordinate specialized sub-agents for complex workflows.
RAG Pipelines
Building retrieval-augmented generation systems with hybrid search, vector databases, and intelligent re-ranking for accurate answers.
LLM Infrastructure
Deploying production LLM services on AWS with observability, guardrails, prompt regression testing, and auto-scaling.
Backend Systems
Engineering high-performance APIs and microservices with FastAPI, Spring Boot, Kafka, and distributed caching.
AI Safety & Guardrails
Implementing PII detection, hallucination prevention, off-topic filtering, and prompt injection defense for enterprise AI.
MLOps & LLMOps
End-to-end AI lifecycle with LangSmith observability, MLflow tracking, automated evaluation harnesses, and CI/CD for prompts.
About
Hi, I'm Sai Pavan.
Agentic AI Engineer at Toyota Motor North America with 4+ years of software engineering experience. I specialize in multi-agent orchestration, RAG pipelines, and production LLM deployment.
I transitioned into AI engineering during my MS in Computer Science at Wichita State University (GPA 3.97), applying a strong distributed systems and backend foundation to LLM infrastructure. Previously, I spent 3 years at Sainsbury's building Spring Boot microservices handling high-volume e-commerce transactions.
const SaiPavan = {
role: "Agentic AI Engineer",
company: "Toyota Motor North America",
stack: [
"LangGraph",
"LangChain",
"GPT-4o",
"AWS Bedrock",
],
focus: "Multi-Agent Systems",
status: "Building the future of AI",
};Career
Experience
Agentic AI Engineer · Toyota Motor North America
Jun 2025 – PresentBuilding multi-agent orchestration platform with LangGraph and OpenAI Assistants API for connected-vehicle diagnostics. RAG pipeline with AWS Bedrock + Pinecone. LLM observability with LangSmith.
AI Software Engineer Intern · Toyota Motor North America
Aug 2024 – May 2025Built prototype LLM-powered supply chain assistant. Designed document intelligence pipeline. Built automated agent evaluation harness with pytest + LangSmith.
Software Developer · Sainsbury's
Aug 2020 – Jul 2023Spring Boot microservices for e-commerce platform. Distributed caching layer with Redis. Monolith-to-microservices migration on AWS.
Get in Touch
Let's build something
great together.
Have a project in mind? I'd love to hear about it. Send me a message and I'll get back to you within 24 hours.
saipavank809@gmail.com
Location
Dallas, TX
Availability