Agentic AI Engineer
BALAJI G
AI Engineer building Agentic AI, LLM, and RAG-based systems in production. Experienced with LangChain, LangGraph, AutoGen, ADK, Semantic Kernel, and Microsoft Agent Framework (MAF), with graph-based retrieval using Neo4j and GraphDB.
Bengaluru, India🏆 GDG Chennai 1st PlaceAgentic AIRAG + Graph Retrieval
8
Autonomous agents
34
Skills delivered
92%
Retrieval accuracy
Top 50
Lyzr Agentathon
Projects
01 — Ongoing
Enterprise App Generation Platform (KSpec)
Built a spec-driven pipeline that generates plans and specs for Greenfield and Brownfield enterprise applications before code generation, with LangGraph orchestration across OpenHands agents.
2026
OpenHandsLangGraphSpec-Driven Development
SDD — plan + spec generation
Dual model strategy — cost + latency optimized
02 — Completed
GitHub Copilot Go Lambda Generation Pipeline
Built a 3-skill workflow (parser, mapper, writer) with complexity-based template selection to auto-generate new API modules from extracted Go code patterns.
2026
Copilot CloudGoMulti-Agent
8 agents — autonomous pipeline
34 skills — productionized workflow
03 — Completed
Multi-Agent Banking Conversational System
Built an autonomous banking assistant with adaptive agent routing, voice interaction, and RAG-based policy retrieval for Forex, Loan, and Credit workflows.
2025
AutoGenACPA2ARedisRAG
92% — retrieval accuracy
Voice — STT/TTS enabled
04 — Hackathon Winner
CiviqAI — Civic AI Platform
Built and owned an ADK-based multi-agent pipeline that auto-routes civic complaints to the correct municipal department with Gemini Vision severity classification and real-time dashboard tracking.
Feb 2026
PythonADKFastAPIRedis
1st Prize — GDG Chennai
Real-time — dashboard tracking
Experience
Mar 2025 –
Present
Kumaran Systems Pvt Ltd
Agentic AI Engineer / Generative AI Engineer
- Building an enterprise-grade Greenfield/Brownfield application generation platform using OpenHands agents orchestrated through LangGraph nodes and spec-driven development.
- Designed dynamic LLM routing in the SDD pipeline: lightweight models for coding/execution, high-capability models for planning/spec generation (KSpec).
- Reduced manual Go Lambda API migration effort to near-zero by architecting a GitHub Copilot Cloud pipeline with 8 autonomous agents and 34 skills.
- Shipped conversational banking agents for Forex, Loan, and Credit workflows using AutoGen, ACP, and A2A with autonomous orchestrator routing.
- Integrated RAG, MCP, and vector search for banking policy/regulation retrieval, achieving 92% retrieval accuracy on complex long-context queries.
2024
Research Fox Consulting Pvt Ltd
Software Intern (Data)
- Built a sentiment analysis chatbot on Telegram for 500+ users, automating review collection and reducing manual analysis effort.
- Delivered the automation using Streamlit and WhatsApp.web.js integrations.
Technical Skills
Languages
PythonJavaSQL
AI/ML & GenAI
LLMRAGAgentic AIPrompt EngineeringVector EmbeddingsSemantic SearchMultimodal AI
Agent Frameworks
LangChainLangGraphAutoGenACPA2AADKMCPSemantic KernelMicrosoft Agent Framework (MAF)DeepAgentsOpenAI Agents SDKOpenHands
Retrieval & Graph
RAGVector SearchGraphDBNeo4j
LLM Platforms
OpenAIAzure OpenAIGeminiHugging Face
Backend & APIs
FastAPIFlaskSpring BootREST APIsMicroservices
Data & Cloud
RedisMongoDBMySQLAWS LambdaAzure
Achievements
1st Place, GDG Chennai Hackathon — Built and owned a multi-agent civic AI platform end-to-end (Feb 2026)
Reward and Recognition (R&R) Award (x3), Kumaran Systems — Recognized for consistent high performance
Go Extra Mile Award, Kumaran Systems — Recognized by Senior VP for outstanding ownership and delivery
Top 50, Lyzr Agentathon — out of 500+ participants
Recognized as Best Team Player for collaborative delivery across cross-functional projects
Education & Certifications
2020 – 2025
Vellore Institute of Technology
M.Tech, Computer Science Engineering — GPA: 8.11/10
Multi-Agent Orchestration — Autogen ACP Framework, DeepLearning.AI (2025)
Certified Agentic AI Engineering — Udemy (2025)
Spring Boot Masterclass — Udemy (2025)
MongoDB — Building RAG and AI Agents
Let's build something.
Open to AI Engineer roles focused on Agentic AI, LLM systems, RAG, and scalable backend platforms.