Speaker : Dr. Andrew Maas, VP of AI at micro1
Title : Lessons from Building AI Companies Before and During the LLM Era
Date : 2026-05-14
Time: 13:00 - 14:00
Location : Fotakis Meeting Room - FORTH
Host : Panagiotis Tsakalides
Abstract:
Building successful AI companies requires navigating through uncertainty toward a moving goal. The arrival of large language models (LLMs) and foundation models has shifted the ground beneath product teams more dramatically than any prior AI technology wave. In this talk, I will share lessons from founding and advising several machine learning startups, including the journey from early-stage research prototypes to deployed products serving real users. I will also provide forward-looking perspectives on how recent AI advances impact the evaluation of new business plans, the feasibility of AI-powered products, and the development cost of creating new systems. The talk is intended to be useful both for researchers considering commercialization and for those who want a clearer picture of what's actually happening inside today's AI companies. I will share company case studies, past and present, and leave time for open discussion.
Bio:
Andrew Maas is VP of AI at micro1, where he leads the development of data-centric ML systems that produce training data, evaluations, and expert feedback for frontier AI labs. He is also the founder of DeepForest.org, which operates living laboratories to explore how robotics and AI can scale up carbon-negative land management practices.
Andrew completed his PhD in Computer Science at Stanford in 2015, advised by Andrew Ng and Dan Jurafsky, with dissertation work on deep neural networks in speech recognition. His research was supported by a National Science Foundation Graduate Research Fellowship. He holds a BS in Computer Science and Cognitive Science from Carnegie Mellon University (2009). At Stanford, he continues to teach CS 224S: Spoken Language Processing and supports AI efforts within the StartX founder network.
His industry career spans several machine learning startups and research organizations. He was co-founder and Chief Scientist of Roam Analytics, a healthcare natural-language-processing platform acquired by Parexel, and a research scientist at Semantic Machines, which was acquired by Microsoft. He went on to lead teams as Senior Manager in Apple's Special Projects Group, building data-centric deep learning approaches and tooling for robot autonomy. In 2023, he founded and led Pointable as CEO, developing optimized retrieval systems for enterprise LLM applications, until its acquisition by Apple in 2025. Following the acquisition, he served as Director of Engineering at Apple, working on deep learning architectures for speech technology and LLM-based approaches to spoken dialogue systems.