DetectSynth · try it live — real-time synthetic-voice detection for telephony.

Streams 500 ms frames over WebSockets to a CPU-only FastAPI server at p99 151 ms against a 500 ms real-time budget. Fusing two 242K-param LFCC-LCNNs cut EER 24.90% → 20.26% across 13 attacks never seen in training — +4.64 points, 95% paired-bootstrap CI [4.26, 4.99], over 71,237 utterances.

The two models fail on disjoint attacks, which is the reason fusion works rather than a happy accident. I report the eval number, not the flattering dev one — dev shares attacks with training. Root-caused a SIGSEGV under load to shared mutable model state and took verified concurrency from 3 → 20 sessions. 156 tests across 17 suites.

PythonPyTorchFastAPIWebSocketsDockerNext.js

SQLGym — a containerized OpenEnv-spec (Meta PyTorch × Hugging Face) RL environment that grades LLM data-analyst agents on SQL business questions. Selected top ~3% of 72,000+ at the Meta PyTorch OpenEnv × Scaler hackathon.

Then I audited it after it was selected. The validator built the container but never ran it — so the served API silently failed every query (SIGALRM is main-thread-only; FastAPI runs sync endpoints in a threadpool). I also found an agent reward-hacking path: REPLACE/ATTACH slipped past the regex denylist and could rewrite the data so a wrong answer graded correct.

Both are now closed at the layer that owns the guarantee — a SQLite authorizer allowlist instead of a regex, and ground truth derived from each task's own reference SQL so it cannot drift. Pinned by 33 tests and CI that boots the container and asserts over HTTP.

_{Reward per action, read-only enforcement caught by SQLite itself, and seeded replay. No API key, no cherry-picking — python demo.py, reproducible by anyone.}

PythonFastAPIDockerSQLitePydanticGitHub Actions

Orchestrate Agent — autonomous support-triage agent on an 8-layer DAG, built for the HackerRank Orchestrate Hackathon. Hybrid retrieval (vector + BM25) feeds an LLM, then a second adversarial-critic pass blocks answers that aren't traceable to the corpus → 93% groundedness across 772 documents.

Veloce-AI — chat with any YouTube video in real time. RAG over Pinecone with Groq Whisper-v3 and Llama-3.3-70B. Beat Groq's 25 MB upload cap by extracting 32 kbps audio-only streams (~60% smaller), which unlocked hour-long videos; metadata filtering guarantees zero cross-talk between sessions.

PythonRAGChromaDBBM25PineconeLlama-3.3FastAPI

MutualMind(live) — full-stack AI mutual-fund advisor, ~19.7K LOC across 52 REST endpoints, pairing a computed financial-analytics engine with an LLM advisory layer.

XIRR via Newton–Raphson root-finding and a 500-run Monte-Carlo simulator for return projection, pinned by an 83-test suite covering numerical edge cases. The chatbot uses vectorless RAG — a two-hop LLM walk over a 39-section knowledge tree, no vector DB — which eliminated hallucinated tax rates rather than prompting around them. Hardened auth (boot-time JWT validation, IDOR-safe) and took dependency CVEs from 16 → 3.

Currently AI Engineer Intern @ goAI Solutions, owning multi-tenant model routing end-to-end — policy-based routing, persistent config, and production fallback paths for graceful degradation under upstream failure.

NodeExpressMongoDBReactGeminiFastAPIPostgreSQL