AVAILABLE NOW
Los Angeles · open to relocation

Swaminathan Chellappa

AI engineer who builds and deploys LLM systems into production

RAG pipelines, LLM guardrails, and evaluation frameworks at USC's DILL Lab, including a HIPAA-compliant annotation platform now adopted by an external research organization. Previously shipped features to 23M+ users at Wells Fargo. MS Computer Science, USC '26.

Email meResumeGitHubLinkedInWrite-upsPublished Research

SECTOR 1

Projects

PII Redactor for Gemini

Privacy-first Chrome extension running Llama-3-8B locally in-browser: 98% PII detection at ~2s inference, with automatic rehydration of redacted terms in responses.

TypeScript
Llama-3-8B
Browser APIs
WXT
Chrome Extension

Providentia

Sprint intelligence dashboard unifying Jira, GitHub, CI, and calendar data. A rule engine detects issues across seven categories, scores them, and assigns a criticality tier; an optional AI layer adds a plain-language summary and suggested action per issue, with every sentence verified against source data before it's shown.

Next.js
Mantine
Rule engine
Verified LLM

Voice agent with live PII redaction

Real-time STT → LLM → TTS pipeline with an in-stream redaction layer that strips PII from transcripts before they reach the model. Built on LiveKit Agents.

LiveKit
Deepgram
Python

SECTOR 2

Experience

USC DILL Lab logo

Graduate Research Assistant · USC DILL Lab

Dec 2025 – May 2026

  • "Live Codebook" system turning annotator corrections into LLM prompt rules: F1 up ~30% over zero-shot, annotation time from 3 hours to 12 minutes
  • HIPAA-compliant local anonymization (87% entity detection); deployed across 4 GPUs on USC's HPC cluster, adopted by an external org
koderAI logo

AI software engineering intern · koderAI

Jun – Aug 2025

  • LLM benchmarking and routing framework across 32 models: production accuracy up ~26%, API retries down ~35%
  • Guardrail endpoints (content + model-based filtering) securing internal AI workflows against prompt injection
Wells Fargo logo

Software engineer · Wells Fargo

Aug 2022 – Jun 2024

  • Shipped BillPay and Alerts modules on the mobile platform serving 23M+ users; 100% WCAG accessibility compliance
  • Sentiment dashboard over 10M+ app reviews, contributing to the app's 4.6 → 4.8 star rating
DataWeave logo

Machine learning intern · DataWeave

Nov 2021 – Jul 2022

  • AWS S3 + MySQL inference pipeline with caching and batch orchestration: latency down 98% at ~100K records/day

PIT WALL

Skills

AI/ML

LangChainLangGraphRAGPyTorchHuggingFaceOllamaFAISSLLM eval

Languages & frameworks

PythonTypeScriptReactNext.jsNode.jsFastAPIKotlin

Infra & data

DockerAWSGCPSLURMMongoDBMySQLChromaDB

Testing & quality

TDDJestJUnitEspressoMockitoMockk