Sai Ram Muthineni
AI Engineer & Security ExpertOSCP-certified AI engineer specializing in LLM security, multi-agent systems, and production AI infrastructure. Building LLM gateways, autonomous red team platforms, coding agents, MCP servers, and observability platforms used by real users. 9+ projects shipped, 5 live products, 100% security benchmark scores.
Technical Skills
Experience
- Web application penetration tests and vulnerability assessments
- Identified and reported critical security vulnerabilities
Certifications
Contact
Projects 8 shipped
Autonomous AI red team platform. 5-node LangGraph attack pipeline with LLM-as-Judge evaluation. Learns from every successful attack.
Zero-trust MCP governance gateway. 100% benchmark scores on prompt injection, JWT, OPA bypass, and PII redaction. Agentic SOC runs 24/7.
Production LLM security gateway at brektra.com. Drop-in reverse proxy for OpenAI/Anthropic SDKs.
Autonomous 8-node coding agent. Plans, writes, tests, debugs, and commits code with zero human intervention.
MCP protocol server enabling LLMs to interact with GitHub via natural language. Claude Desktop & Codex CLI compatible.
LLM observability platform with JWT auth, multi-tenancy, trace replay. Published PyPI SDK.
Lightweight LLM observability with Streamlit dashboard. Deployed on GCP.
SaaS automating pentest report generation with LLMs. Real users since launch.
Open Source
PR #3526 merged — Cohere ClientV2.embed instrumentation. PR #3504 approved — Anthropic usage.input_tokens null guard.
PR #4942 merged — 3 fixes squashed: GitHub stargazer fixture tests, stale CLI test rename, goal module docstring cleanup.
PR #4777 — yaml.BaseLoader fix for skills frontmatter parsing.