current
Partner Engineer, Applied AI
@ LangChain
May 2026 — Present|New York, NY
$ apt list --installed
LangChainLangGraphDeep AgentsLangSmithNVIDIA NemotronAWS BedrockMCP ProtocolPythonTypeScript
metrics.sh
$ metrics --role=langchain-partner-engineer
> Flagship Launch: NemoClaw Deep Agents Blueprint (with NVIDIA)
> Benchmark Result: Benchmark-leading open-model agents at ~10x lower inference cost
> Merged Oss Prs: 10+ across LangChain public repos
> Partners Enabled: 15+ across the AWS, GCP & Azure ecosystems
> Published Content: 6+ blogs & cookbooks, 8+ talks and workshops
Partner Engineer on the Applied AI team — the technical bridge between LangChain and strategic ecosystem partners like NVIDIA, AWS, and Google Cloud, turning partnerships into shipped integrations, benchmarks, joint solutions, and go-to-market assets.
Highlights
- •Led LangChain's technical engagement with NVIDIA — tuned and benchmarked LangChain Deep Agents on the Nemotron 3 Ultra (550B) open model and authored the open-source harness profile (tool-call shims, compatibility and execution-repair middleware, loop control) behind the jointly launched NemoClaw Deep Agents Blueprint: benchmark-leading agent performance among open models at ~10x lower inference cost than leading closed models
- •Drove LangSmith's launch on the AWS Marketplace — took the SaaS listing from private to public, completed the AWS Foundational Technical Review (FTR) to unlock co-sell, and authored reference architectures and enablement assets for AWS field teams
- •Built self-serve tooling adopted across the GTM org — a LangSmith cloud cost calculator and a Workshop Studio that auto-provisions temporary enterprise trial environments for workshop attendees
- •Helped launch LangChain's partner certification program and the LangChain Certified Agent Engineer credential, co-authoring exam content and the certification curriculum
- •Delivered technical enablement to 15+ partners and SIs across the AWS, GCP, and Azure ecosystems — POCs, evaluations, and hands-on workshops on LangGraph, Deep Agents, and LangSmith
Open Source
10+ merged pull requests across LangChain's public repositories under srimanthtangedipalli-eng:
- •NVIDIA Nemotron 3 Ultra harness profile — deepagents #4192 — the flagship contribution behind the NemoClaw blueprint (tool-call shims plus compatibility and execution-repair middleware)
- •Deep Agents AWS tour — six merged PRs of sample notebooks and hands-on labs in LangChain's AWS samples repo
- •OpenShell contained-agent workshop in LangChain's NVIDIA samples repo
- •Documentation fixes and co-authorship of the Crusoe chat-model integration
Writing
Posts I've authored on the LangChain Blog:
- •Deep Agents Code on NemoClaw: a governed blueprint for your most sensitive code — Jul 2026
- •Tuning the harness, not the model: a Nemotron 3 Ultra playbook — Jul 2026, with Nick Hollon
- •Building an auditable VC research agent with the Perplexity Agent API and LangGraph — Jun 2026, with Karan Singh and Ryan Buchmayer
- •Building a company due diligence agent with Deep Agents, LangSmith and Parallel — May 2026, with Matt Harris, Nick Martitsch, and Karan Singh
The Nemotron harness work has also been featured on the NVIDIA Developer Blog and the NVIDIA Blog, and syndicated by inference partners including Fireworks, Crusoe, Nebius, DeepInfra, and Baseten.