Trinath Manikanta B

About Trinath Manikanta B

I build AI systems that actually ship to production.As an AI/ML Engineer with 4+ years of experience, I specialize in translating cutting-edge research into scalable, real-world applications — from LLM fine-tuning and agentic reasoning systems to multimodal generative AI pipelines across AWS, Azure, and GCP.At Dexcom, I\'m driving GenAI adoption at the enterprise level:• Fine-tuning GPT, Claude & LLaMA using SFT, LoRA/PEFT — boosting model accuracy by 18% and F1 by 14% Architecting RAG systems with FAISS & Pinecone for enterprise knowledge retrieval• Engineering LangChain/LangGraph agentic frameworks for autonomous analytical workflows Cutting inference latency by 32% via scalable Kafka/Spark pipelines and REST microservicesPreviously at Virtusa (Citi Bank), I built production ML solutions for credit-risk modeling and fraud detection — improving fraud detection accuracy by 21% and reducing fraud by 17% on large-scale financial datasets. Research-to-Production Specialties:• Multimodal Vision-Language Models (CLIP, LLaMA, RLHF/DPO) Generative AI: Diffusion Models, LoRA personalization, Genmoji-style content generation• Privacy-Preserving ML & Responsible AI governance MLOps: MLflow, RAGAS, DeepEval, Prometheus, CI/CD, Docker, Kubernetes Currently based in Portland, Oregon | Open to Senior AI/ML Engineer & Applied Scientist roles Let\'s connect: t••••••••@gmail.com

Experience

1. Ai Ml Engineer

   Dexcom

   Jun 2025 — Present · Portland, OR, US

   Pioneered GenAI applications using GPT, Claude & LangChain/LangGraph; fine-tuned LLMs (SFT, LoRA/PEFT) and architected FAISS/Pinecone RAG systems — improving retrieval accuracy by 18% and F1 score by 14%. Built agentic reasoning frameworks (LangGraph, AutoGen) enabling autonomous multi-step analytical workflows, reducing manual analyst workload by 40%.• Engineered Kafka + Spark/PySpark data pipelines with REST microservices for real-time AI workloads — cutting inference latency by 32% across critical enterprise platforms. Instrumented end-to-end MLOps with MLflow, RAGAS, Prometheus & Grafana; maintained responsible AI standards, regulatory compliance, and governance documentation.

Education

• INSTITUTE OF AERONAUTICAL ENGINEERING

  Bachelor of Technology - BTech, Electrical, Electronics and Communications Engineering

  2018 — 2022

• University of South Florida

  Master's degree, Data Intelligence