I'm Sumanth Polisetty, a Senior AI & Data Science Professional with 7+ years of experience building enterprise-scale GenAI, Multi-Agentic Workflows, and MLOps solutions.
Agentic RAG, Multi-Agentic Systems, LangGraph, Model Context Protocol (MCP), AWS Bedrock, OpenAI, Claude.
AWS (SageMaker, Lambda, OpenSearch), Azure (AKS, AI Foundry, Cognitive Search), Docker, Kubernetes.
Python, Scikit-learn, Transformers, PyTorch, TensorFlow, Hugging Face, FAISS, ChromaDB.
An enterprise-grade open-source platform for assessing AI engineering skills. It simulates "broken production" scenarios, requiring candidates to use LangGraph, MCP, and RAG to resolve issues in a sandboxed environment.
def assess_candidate(skills):
container = spawn_sandbox()
run_llm_judge(container)
return "HIRE"
> Containerizing code-server IDE...
> Initializing Ollama Judge...
> Ready for evaluation.Stay tuned for more Agentic architectures and enterprise AI tools.
Architected multi-agent workflows using LangGraph and AWS Bedrock, reducing month-end financial close times by 90%. Engineered an enterprise-grade MCP server for secure financial operations across SAP, Coupa, and Tipalti.
Led GenAI COE initiatives, building advanced Multi-Agentic and RAG solutions for Manufacturing and Mortgage platforms. Deployed MLOps pipelines managing production models on Azure Kubernetes Service (AKS).
Developed demand forecasting models with temporal fusion transformers and implemented anomaly detection on IoT sensor data. Built semantic search engines using Azure Cognitive Search.
Designed a Unified Identity Platform using OAuth 2.0 that centralized authentication for diverse consumer products. Optimized REST API performance by 25% with AWS Lambda-based serverless functions.
I write about practical lessons from building real-world GenAI systems, exploring topics like the hidden costs of AI coding tools, building AI operators with OpenClaw, and the nuances of Agentic architecture.
Managing "AI Slop" in Real Systems without accumulating technical debt.
Read Article →A hands-on guide to setting up OpenClaw locally on an RTX Laptop.
Read Article →The engineering behind Applied AI and understanding vector space.
Read Article →Exploring the future of OpenClaw and Enterprise AI Operators.
Read Article →