AI/ML · Cybersecurity · Blockchain
Cybersecurity Analyst specializing in AI/ML-enhanced security and blockchain auditing. I build intelligent systems that detect threats, audit smart contracts, and automate security workflows. Currently based in Pennsylvania.
MS Cybersecurity
Gannon University · 2024–2025
BTech AI/ML
Amrita Vishwa Vidyapeetham · 2019–2023
Work
Live product In developmentNo dot — completed
Live Products
Educational Next.js + Solidity lab: flash-borrow demo capital, spike an unsecured spot oracle, VulnerableVault drain succeeds in-lab vs MitigatedVault (TWAP + circuit breaker) blocks it. Evidence 2026-09-15: npx hardhat test — 6 passing. Demo funds only — not an exploit kit or mainnet toolkit.
SOC-style SIEM lab dashboard (Next.js). Real TypeScript detectors for SSH brute force (T1110.001), SQLi (T1190), and port scan (T1046) on RFC 5737 demo logs, plus a deterministic pattern hunter at /hunt (no LLM/threat-intel APIs). Evidence: deployed at buggerhunt.vercel.app with Railway Postgres; 29 vitest tests. Lab/demo — not Elastic, not a production SOC.
FastAPI lab: local PDF/CSV/Excel text extract, then optional Groq Llama 3.3 structuring. Evidence 2026-09-09: imports + CSV extract proven. No OCR. Live extract UI on Railway (api_key_set may be false = extract-only). https://documind-free-production.up.railway.app
Python CLI that splits each Solidity function and flags reentrancy-style calls, missing owner checks, unchecked low-level calls, block.timestamp, and tx.origin. Optional Slither. Sample run 2026-09-09: 7 findings vs 0 clean. Pattern checks only — not ML and not an audit.
Solidity staking vault with ReentrancyGuard and CEI. Evidence 2026-09-09: npx hardhat test — 6 passing, 0 failing. Lab contract — not a mainnet deploy.
Lab ERC-20 where staking rewards and holder dividends compete for the same balance. Evidence 2026-09-09: npx hardhat test — 33 passing. Not a product or audit.
Lab FastAPI + Claude tool-use document agent (read/split/validate/export tools in code). Live lab UI on Railway with Claude tool-use when key configured. Not a production multi-tenant product. https://luminous-compassion-production-1f80.up.railway.app
Jupyter/Colab lab: U-Net + EfficientNetB3 iris vessel segmentation notebook. Educational CV experiment — not a production biometric system.
Completed Projects
Student internship agent (find → local pattern tailor → review-before-submit). Public repo is the product README; implementation stays private. Evidence: pattern tailor without API keys; Fortune 100 career-page tracker seed (Wikipedia top 100; official 1–500 paywalled). No LinkedIn scrape, no password harvest, no auto-submit by default.
Detection rules on generated auth, Apache, and iptables logs (no Elastic/Kibana). Evidence 2026-09-09: 3 alerts — SSH brute force T1110.001, SQLi T1190, port scan T1046. Local stdlib dashboard optional.
Python lab scanner: TCP connect, banners, simple web checks, headers, TLS peek. Evidence 2026-09-09 on scanme.nmap.org: 4 open ports, 5 missing headers. No CVE database — not a pentest.
Under Development
Next.js app that turns a YouTube URL into LinkedIn post drafts (Supabase/Stripe/LLM wired in code). npm build verified 2026-09-09. www.repostly.org currently has an expired TLS cert — do not treat as a reliable live product or as having paying customers.
Docs-only learning stub about Coinbase x402 / HTTP 402. No runnable backend or frontend in the repo (as of 2026-09-09).
Healthcare multi-agent experiment stub. Verify README/code before citing specific agent workflows.
Expertise
AI / ML
Cybersecurity
Blockchain
Development
Get in touch
Open to full-time roles in cybersecurity, AI/ML, and blockchain. Also available for freelance security audits and AI consulting.