Machine Learning Engineer · brain.co

Amit Namburi

ML for whatever it takes

multimodal audio/video·post-training·inference

B.S. Computer Science, UC San Diego M.S. Computer Science, UC San Diego
Resume Get in touch
Portrait of Amit Namburi

About

Machine Learning Engineer at brain.co. I work across the stack: multimodal image/audio models, post-training (SFT, preference optimization), and inference (serving, latency budgets). A growing part of my work is the data layer: large-scale annotation guidelines and labeling pipelines with quality controls, inter-annotator agreement, and schema versioning for consistent, high-signal training and eval data. And I build classical computer vision that powers automated permitting systems deployed in government today.

BS/MS in Computer Science from UC San Diego, June 2026, then straight into startups. In the McAuley Lab I worked on audio-language pretraining, music reasoning, and conversational recommendation, with papers at ICASSP, EMNLP, and ISMIR. I also TA'd recommender systems and ML for music.

Love building in fast-paced environments.

Experience

Machine Learning Engineer

brain.co · San Francisco, CA

Jul 2026 – Present

  • Multimodal model development, post-training (SFT, preference optimization), and low-latency inference for production systems.
  • Classical computer vision and full-stack work to ship and maintain a permitting application used in government today.

Member of Technical Staff, Machine Learning

NomadicML · San Francisco, CA · Remote

Oct 2025 – Jun 2026

  • Built vision-language-action (VLA) models for efficient large-scale video dataset search and understanding.
  • Efficient long-tail edge case detection for autonomous driving.
  • Product used by Amazon Zoox, Qualcomm, and more.

Software Engineer Intern

Apple · CoreOS · San Diego, CA

Jun 2025 – Sep 2025

  • Built E2E SoC regression detection from performance telemetry.
  • Partnered with Siri on an LLM-as-a-judge evaluation framework for audio tasks.
  • iContest finalist, Apple's internal intern innovation competition.

UC San Diego

San Diego, CA · Apr 2023 – Jun 2026

Graduate Student Researcher

McAuley Lab · Jan 2024 – Jun 2026

PI · Julian McAuley

  • Multimodal LLMs for audio and visual understanding; long-form audio-language pretraining.
  • Fine-tuning and evaluation methods for multimodal tasks and conversational recommendation.

Teaching Assistant

CSE Department · Mar 2025 – Jun 2026

  • CSE 158/258: Recommender Systems & Web Mining (Fall 2025).
  • CSE 153/253: Machine Learning for Music (Spring 2025).

Instructional Assistant

CSE Department · Apr 2023 – Mar 2025

  • Tutored CSE 100 (Advanced Data Structures) and CSE 12 (Data Structures & OOD).
  • Delivered 800+ hours of tutoring, mentoring 1,000+ students through office hours and review sessions.

Computational Researcher

FAIR Data Informatics Lab · San Diego, CA

Mar 2023 – Mar 2025

PI · Anita Bandrowski

  • Redesigned the Foundry Dashboard, an ETL system for biomedical data curators.
  • Built a RAG pipeline with LlamaParse, LlamaIndex, and NVIDIA NIM microservices.

Vice President

UCSD Computer Science & Engineering Society · San Diego, CA

Apr 2022 – Jun 2023

  • Led student programs and community initiatives; hosted events with CSE faculty.
  • Organized technical workshops and networking events.

Machine Learning Intern

IPMD, Inc. · Part-time · Berkeley, CA · Remote

Jun 2022 – Sep 2022

  • Built and shipped traditional CV models to recognize facial expressions.
  • Designed and implemented scalable APIs to utilize models in real time.

Publications

2026
ICASSP

MusiCRS: Benchmarking Audio-Centric Conversational Recommendation

Rohan Surana*, Amit Namburi, Gagan Mundada*, Abhay Lal*, Zachary Novack, Julian McAuley, Junda Wu

ICASSP 2026 · IEEE Intl. Conference on Acoustics, Speech and Signal Processing

arXiv ↗ Scholar ↗
2026
EMNLP

WildScore: Benchmarking MLLMs In-the-Wild Symbolic Music Reasoning

Gagan Mundada, Yash Vishe, Amit Namburi, Xin Xu, Zachary Novack, Julian McAuley, Junda Wu

EMNLP 2026 · Empirical Methods in Natural Language Processing

arXiv ↗ Scholar ↗
2025
ICASSP

FUTGA-MIR: Fine-grained & Temporally-aware Music Understanding with MIR

Junda Wu, Zachary Novack, Amit Namburi, Hao-Wen Dong, Carol Chen, Jiaheng Dai

ICASSP 2025 · IEEE Intl. Conference on Acoustics, Speech and Signal Processing

IEEE Xplore ↗ Scholar ↗
2025
ICASSPSpotlight

CoLLAP: Contrastive Long-form Language-Audio Pretraining with Musical Temporal Structure Augmentation

Junda Wu, Warren Li, Zachary Novack, Amit Namburi, Carol Chen, Julian McAuley

ICASSP 2025 · IEEE Intl. Conference on Acoustics, Speech and Signal Processing

IEEE Xplore ↗ Scholar ↗
Soon

More in review. In the meantime: Semantic IDs for generative recommendation, and some classical computer vision.

Follow on Scholar for updates.

Contact

Emailanamburi@ucsd.edu LinkedIn/in/amit-namburi GitHub@namburiamit ScholarGoogle Scholar ResumePDF