Reeth Jain
ML Engineer | GenAI & Agentic AI Systems | MLOps
Building production-grade AI systems with focus on Large Language Models, Retrieval-Augmented Generation, and autonomous agent frameworks. Experienced in deploying scalable ML infrastructure from research to production.
Professional Experience
AI Developer Intern
Hello agentic.ai
- ▹Designed and developed agentic AI systems with LangGraph state-machine workflows, enabling autonomous task decomposition, planning, and coordinated multi-agent execution
- ▹Built robust tool-augmented pipelines integrating memory, retrieval, and external APIs so agents could handle complex multi-step tasks with better context retention and decision quality
- ▹Implemented production-focused reliability patterns including structured error handling, retries, guardrails, and fallback paths to improve safety and reduce workflow failure rates
- ▹Collaborated on deployment readiness by improving observability, evaluation loops, and maintainable Python module design for scalable agent operations
AI SDE Intern
Xelron.ai
- ▹Built 15+ production-grade AI agent benchmarks using Dockerized environments, pytest suites, and TOML-driven configs to ensure reproducible Harbor evaluations across runs
- ▹Engineered an automated validation harness with 11 quality gates covering format checks, deterministic outputs, edge-case handling, and scoring consistency to reduce benchmark drift
- ▹Standardized task specifications with JSON schemas, explicit constraints, and adversarial edge-case definitions, significantly improving reproducibility, traceability, and auditability
- ▹Delivered benchmark packs across ML, data processing, security, and scientific-computing domains, expanding evaluation coverage and enabling more reliable cross-domain model comparisons
Featured Projects
Autonomous Security MLOps Platform
Enterprise-grade MLOps platform with 5 ensemble models, advanced monitoring, drift detection, and production-safe deployment infrastructure.
Hallucination-Aware Hybrid LLM System
Production-grade RAG system with Phi-3, FAISS retrieval, and strict context-grounding to prevent hallucinations. 86.7% retrieval accuracy.
VibeTune: Multi-Modal Emotion Music Recommender
Real-time emotion detection from face, voice, and text with Spotify integration. ResNet50 + Wav2Vec2 + DistilRoBERTa for multi-modal analysis.
GenAI YouTube & Web Summarizer
AI-powered summarization tool using Whisper for transcription and LLMs for intelligent content summarization with 5 summary styles.
Technical Expertise
AI/ML Frameworks & Libraries
LLM & GenAI Tools
Backend & APIs
DevOps & Cloud Infrastructure
MLOps & Observability
Development & Analysis
Let's Build Something Impactful
Open to ML Engineering, GenAI, and MLOps roles where I can contribute to production-grade AI systems and scalable infrastructure.
"Building AI systems that scale from prototype to production"