I'm an independent security researcher, developer.
I build tools that catch things scanners miss, research vulnerabilities before they become headlines, and occasionally write papers about obscure attack surfaces like maritime AIS spoofing and LLM trust chains.
Right now I'm deep into:
- Supply chain security — behavioral analysis, syscall tracing, ML anomaly detection
- AI/LLM security — API repackaging attacks, MCP server trust boundaries, agent threat modeling
- Bug bounty — active engagements, building automation tooling around the hunt
|
Runtime behavioral analysis tool that maps the process cascade of suspicious packages into a directed tree — catching supply chain attacks that install-time scanners miss. Eight detection "legs": sandbox isolation, syscall parsing, behavioral graphing, ML anomaly detection (Random Forest, 841 samples), YARA signatures, MCP security protocol, and more. |
An MCP server that transparently routes AI coding agent requests across multiple models and providers. Never hit a rate limit dead end again. |
|
Empirical validation of the Mirror API Effect — quantifying 8051% latency penalties from API repackaging and proposing Zero-Trust UIC architectures for the digital supply chain. Companion to my SSRN paper on LLM trust vulnerabilities. |
A fully local, privacy-first AI Text-to-Speech studio and podcast creator. Powered by Kokoro TTS + RVC with a cinematic neuromorphic UI and emotion-aware narration. Your voice, your machine, no cloud. |
| Paper | Venue / Status | Topic |
|---|---|---|
| Mirror Effect: API Repackaging & LLM Trust Vulnerabilities | SSRN (preprint) | AI security, supply chain |
| MoE TTS Routing Architecture | Zenodo | ML infrastructure |
| The Cascading Visibility Failure in Maritime Logistics | Zenodo | Maritime AIS spoofing, cyber-physical security |
- LinkedIn: tejas-prasad-a8b891334
- X / Twitter: @Tishieboi)
- ORCID: 0009-0008-7201-7799
- Scholar: Google Scholar
"Just pushing all my work in here, because its worth being published... I think."