Nikunj Kotecha
Senior Machine Learning Engineer | Edge AI & LLM Specialist | AWS Certified | Quantization, RAG, LoRA Expert
- Role
- Senior Machine Learning Engineer at Coastaloupe
- Location
- Irvine, CA, US
- LinkedIn followers
- 500 followers
About Nikunj Kotecha
Cutting AI Latency, Driving Business Wins, and Bringing Models to the Edge.In today’s AI race, companies face a common challenge: models that are too slow, too power-hungry, or too inaccurate for real-world deployment. That’s where I come in.I am Nikunj Kotecha, a Senior Machine Learning Engineer & AWS Certified AI Practitioner with 8+ years of experience turning advanced ML research into revenue-generating, hardware-ready solutions. My career spans from building neuromorphic PoCs at BrainChip that secured $2M+ in global IP wins, to deploying RAG-powered clinical assistants that cut misdiagnosis rates in low-connectivity clinics by 27%.My specialty: bridging the gap between cutting-edge architectures (LLMs, LoRA, SSMs, multimodal transformers) and the constraints of edge devices. I’ve reduced inference latency by up to 90%, shrunk models by 75% without sacrificing accuracy, and built retraining pipelines that keep AI relevant in production.Along the way, I have led technical teams, founded the BrainChip University AI Accelerator program, lectured at Carnegie Mellon, ASU, and RIT, and co-authored peer-reviewed publications in and ACM. Whether it is optimizing a 4-bit quantized object detection model or designing a hardware-aware transformer, I focus on one thing: making AI faster, smarter, and more deployable.Off the clock, I am an active voice in the AI community with publishing articles, mentoring engineers, and researching next-gen architectures.If your AI is too slow, too costly, or too big to deploy, or if you want to collaborate with me, let’s talk. DM me here.
Experience
Senior Machine Learning Engineer
Jan 2024 — Present
Fine-tuned small LLMs with LoRA & PEFT, integrating RAG to boost clinical recommendation relevance- Built quantized, pruned, and distilled models for low-power devices, cutting inference latency- Designed SageMaker-integrated retraining & shadow evaluation pipelines for continuous pre-production validation
Education
University of Mumbai
Bachelor of Engineering - BE, Computer Science
Rochester Institute of Technology
Master of Science - MS, Computer Science & Artificial Intelligence
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.