Location: Gurugram, India (IST)
Remote: Yes (comfortable with US/EU working-hours overlap for standups, reviews, and fast back-and-forth)
Willing to relocate: Yes for the right in-person role (US/Canada/Europe etc.)
Technologies: AI agents, AI security, policy enforcement, supervised runtimes, Python, TypeScript, Linux, Docker, AWS, Java, RAG, tool orchestration, LLM evals, prompt injection / jailbreak testing, reverse engineering
Résumé/CV: https://raw.githubusercontent.com/maniotrix/public-drive/mai... · https://github.com/intentframe/intentframe · https://github.com/maniotrix
Email: [email protected]
---
I built IntentFrame (https://github.com/intentframe/intentframe), an open-source framework for validating AI agent actions before they run. No tool call hits files, credentials, network, APIs, or money without going through policy checks first. Shipped runtime packages to PyPI, integrated it into a production agent stack, and designed it around prevention, not post-hoc monitoring.
That work sits on top of 8+ years shipping production AI systems and Python services for US/EU clients (Arc.dev, long freelance history, 5.0 on Codementor across 93+ reviews). Overlapping on US hours for standups and reviews has been normal for me for years. I also do scoped pre-launch security reviews for mobile + backend apps with written Critical/High/Medium findings.
Before IntentFrame I built KarnAGT, a reasoning agent with RAG and tool orchestration (open source), and earlier reverse-engineered an obfuscated Java client end to end with a custom captcha OCR pipeline running in production 24/7. IIT BHU, Google Summer of Code alum.
Looking for early-stage / founding engineer or senior AI roles where agent security, agent infra, or hard agent products matter. Small team, real ownership. Happy to talk.
Remote: Yes (comfortable with US/EU working-hours overlap for standups, reviews, and fast back-and-forth)
Willing to relocate: Yes for the right in-person role (US/Canada/Europe etc.)
Technologies: AI agents, AI security, policy enforcement, supervised runtimes, Python, TypeScript, Linux, Docker, AWS, Java, RAG, tool orchestration, LLM evals, prompt injection / jailbreak testing, reverse engineering
Résumé/CV: https://raw.githubusercontent.com/maniotrix/public-drive/mai... · https://github.com/intentframe/intentframe · https://github.com/maniotrix
Email: [email protected]
---
I built IntentFrame (https://github.com/intentframe/intentframe), an open-source framework for validating AI agent actions before they run. No tool call hits files, credentials, network, APIs, or money without going through policy checks first. Shipped runtime packages to PyPI, integrated it into a production agent stack, and designed it around prevention, not post-hoc monitoring.
That work sits on top of 8+ years shipping production AI systems and Python services for US/EU clients (Arc.dev, long freelance history, 5.0 on Codementor across 93+ reviews). Overlapping on US hours for standups and reviews has been normal for me for years. I also do scoped pre-launch security reviews for mobile + backend apps with written Critical/High/Medium findings.
Before IntentFrame I built KarnAGT, a reasoning agent with RAG and tool orchestration (open source), and earlier reverse-engineered an obfuscated Java client end to end with a custom captcha OCR pipeline running in production 24/7. IIT BHU, Google Summer of Code alum.
Looking for early-stage / founding engineer or senior AI roles where agent security, agent infra, or hard agent products matter. Small team, real ownership. Happy to talk.
As team lead, built production agentic platform (FastAPI, React, LlamaIndex, sandboxed code execution) and IntentFrame (open-source agentic security runtime with policy enforcement, PyPI).
8+ years delivering Python services, Linux production workflows, and client-facing AI platforms (Arc.dev/Codementor, 93+ reviews). IIT BHU (B.Tech) and Google Summer of Code(GSOC) alum.
Combines hands-on agent engineering with adversarial testing and clear technical reporting for engineering and product teams.
Core Competencies * AI Agent Engineering & Production Systems - Multi-turn agents · tool orchestration & dynamic tool registry · RAG (LlamaIndex, vector search) · planning loops · streaming execution · agent-agnostic integrations (LangChain, AutoGen, OpenAI Agents SDK, MCP) · FastAPI · React/TypeScript · Docker · AWS · production debugging · PyPI packaging * LLM Applications & Full-Stack Delivery - Python services · API design & third-party integrations · auth/session flows · multi-tenant backends · mobile & web (React Native, Flutter, Android) · real-time systems · Linux environments · automation & test scripting * AI Security & Reliability - Policy enforcement · action validation · prompt-injection defense · agent tool misuse testing · red-team scenarios · pre-launch security reviews · fail-closed runtime design * ML, Research & Deep Systems - Adversarial ML · TensorFlow · reverse engineering · Java/Kotlin · bytecode & reflection analysis · attack-surface mapping · root-cause analysis · structured remediation reporting
GitHub: https://github.com/maniotrix IntentFrame: https://github.com/intentframe/intentframe Portfolio: https://prince-ai-architect.vercel.app/ Contact Info: [email protected]