About Me

My work focuses on post-training, data selection, domain adaptation, and personalization. I currently work on foundation models for personalization at Netflix. Previously, I post-trained Microsoft's Phi-3 language models, built generative AI systems at Microsoft, and conducted research at Google Brain and Google AI. I received my Ph.D. in Computer Science from Stanford University, where I was advised by Prof. Dan Jurafsky and was a member of the Stanford NLP Group and Stanford AI Lab.

In parallel with my work at the frontier of AI, I've spent the last decade advising startups, building products, and occasionally founding them. My interests sit at the intersection of AI, company-building, and helping exceptional young founders get off the ground. Some of that lives under Sagur.

Contact Details

Dan Iter
Austin, TX
my full name AT cs DOT stanford.edu

Education

Stanford University

PhD in Computer Science • Present

Ph.D. Thesis: Pre-finetuning Methods for Domain and Task Adaptation, with Applications to Discourse and Translation. My research focused on adapting pretrained language models through an intermediate training stage before task-specific fine-tuning. While we called this pre-finetuning at the time, the approach closely foreshadowed what is now commonly referred to as post-training.

Columbia University

B.S. in Computer Science • May 2011

I graduated Summa Cum Laude. I did research in information retrieval in social data with Prof. Luis Gravano and Hila Becker. My focus was in systems. I was also a member of Bacchanal.

Work

Netflix

Senior Research Scientist • 2026 - Present

Building Netflix's member foundation large language model, the backbone of member-facing personalization, on the Foundation Models / Personalization Foundations team. Research focuses on LLM post-training (DPO, GRPO), distillation including on-policy distillation and inference optimization, data selection, semantic IDs, and model merging to enable next-generation personalization experiences across the Netflix homepage, search, and discovery.

VC-Backed Startup

Member of Technical Staff • 2024 - 2025

Led research in agentic AI and task-specific small language models. Internal research and product development was instrumental in securing over $20 million in investment capital, led by Khosla Ventures. Delivered production models for edge-device and cloud-hosted SLMs, and led the research agenda, production infrastructure, product design, human resource allocation, and mentorship.

Microsoft AI / GenAI

Senior Research Scientist • 2022 - 2024

Member of the Phi-3 post-training team, working on SFT and DPO training of pretrained models for better user alignment, chat functionality, and stronger LLM benchmark performance. Shipped a medical summarization model for Nuance's DAX product that converts doctor-patient conversations into structured EMR summaries. Hosted interns and published papers on efficient finetuning, data selection, and in-context learning for LLMs.

Google Brain

Research Intern/ Student Researcher • June 2020 - Present

Host: David Grangier
Analysis of data selection methods for domain adaptation in machine translation and language modeling. Training transformer-based NMT models using contrastive methods and finetuned-BERT domain classifiers for selecting data most similar to a small target domain.

Google AI Research

Research Intern • June 2019 - September 2019

Host: Kelvin Guu
Coherence objectives for language modeling - learning sentence representation by training transformer language models to understand sentence ordering and discourse distance.

Google Web Answers

Research Intern • June 2018 - September 2018

Entity attribute extraction - a special case of relation extraction where attributes describe searchable aspects of the entity.

Megagon Labs

Research Intern • June 2017 - August 2017

Advisors: Alon Halevy and Wang-Chiew Tan
FrameIt: A system to quickly build framings and SRLs for exploring large text corpora. In progress of building an ontology of happy moments in HappyDB.

Intel

Software Engineer (R&D) • June 2015 - December 2015

I helped design a new highly parallel processor architecture that offers GPU performance on an x86 instruction set. I profiled our design on a number of common machine working workloads.

Infinio

Software Engineer • June 20012 - May 2015

As the 9th member of this startup I was involved from our first prototype to our second major product. I worked across the stack including implemented kernel drivers for high performance distributed caching for datacenters, MVC business logic for system managament and even built our light weight cross platform installer.

-->

Quotes

  • Our opponents maintain that we are confronted with insurmountable ... obstacles, but that may be said of the smallest obstacle if one has no desire to surmount it.

    Theodor Herzl