$ whoami
Backend and AI Infrastructure Engineer.
Seeking Senior Backend, AI Infrastructure, and Platform Engineering roles.
$ status
Building systems that fail loudly instead of silently.
Reading agent traces at 2am to find the one call that shouldn't exist.
I design distributed backend systems and AI infrastructure built to hold up in production.
My work spans agentic AI safety frameworks, fault-tolerant data pipelines processing millions of records, and low-level tooling that stops AI agents from wasting compute on themselves.
I have cut API latency by 40% in a production system and shipped AI safety tooling that runs live rather than staying in a notebook.
Languages Python, TypeScript, Rust
Backend FastAPI, Django, REST, gRPC, GraphQL
AI and Agents LangChain, LangGraph, LLM Ops, vector databases
Data and Streaming PostgreSQL, Redis, Kafka, Elasticsearch
Infrastructure AWS (Lambda, ECS, SQS, S3), Docker, Kubernetes, CI/CD
Practices TDD, observability, fault-tolerant design, performance tuning
Watching how far a local, nanosecond-level check can go before an AI agent's tool call ever hits the network.
Next up: teaching microloop to recognize loop patterns across sessions, not just within one.
AI agents get stuck in infinite tool loops, and each loop cycle burns API credits. microloop detects redundant tool calls locally in nanoseconds and blocks them before the call ever leaves the machine.
Rust. Apache 2.0 license. 4 stars, 1 fork.
Topics: rust, ai, detector, ai-agents, loops-and-iterations, langchain
A Rust systems project focused on stability and correctness under constrained conditions.
Rust. 2 stars.
A behavioral simulation engine for AI safety. It tests agent behavior under stress rather than prompting for compliance.
Email: devaretanmay256@gmail.com
LinkedIn: linkedin.com/in/tanmay-devare
GitHub: github.com/Devaretanmay