
New York, NY, November 18th, 2026, NewsDirect
New Business Unit Delivers a Premium, End-to-End Expert Human Data Offering for AI Training, Evaluation, and RL Environments
Guidepoint, a global pioneer in access to expert insights, today publicly launched its new dedicated Guidepoint Labs team led by human data industry veterans, delivering the full stack of specialized expert data for training and evaluating frontier AI models and agents. This comprehensive offering extends from sourcing and qualifying domain experts to designing tasks, running evaluations, delivering validated data and beyond. Guidepoint Labs, operating in stealth mode for more than six months, is already working with frontier labs, Fortune 100 companies, and AI-native startups, and scaling rapidly to meet demand.
“Frontier labs are increasingly in need of the tacit judgment of experts who are engaged in complex professional work for a living, such as physicians, lawyers, engineers, and financial analysts,” said Albert Sebag, Founder and Chief Executive Officer of Guidepoint. “Guidepoint has spent more than 20 years building a curated pool of industry leaders to service this critical demand for finding enough of the right people at the right time.”
Guidepoint Labs draws on 2M+ Advisors, with 15,000+ specifically recruited each month, across 300+ industries and 150 countries, backed by institutional-grade compliance and vetting. That network enables the team to staff senior and niche expertise quickly and at scale, which generalist annotation vendors can’t match.
“Credentials get an expert in the door. Engagement is what produces training signals a lab can trust,” said Grant Cho, Executive Vice President of Delivery at Guidepoint Labs, formerly with Scale AI and Uber AI Solutions. “Delivering on both criteria has been the hardest problem to solve in my experience, especially for the specialized knowledge and professional judgment that is now needed to continue improving model performance. While Guidepoint’s depth of experts is unmatched, and I’m excited that we’re already seeing 80% of experts willing to engage in these new types of AI training projects. As labs increasingly move into RL environments, a reward is only as good as the expert who defines it.”
Guidepoint Labs provides a wide range of services critical to successful model development and evaluation:
- Training Data: supervised fine-tuning demonstrations, preference data, rewrites, and structured outputs
- RL Environments: domain-specific tasks, tools, and grading rubrics built with practitioners
- Agentic Data: expert trajectories captured in a production-grade harness
- Evaluation and Red Teaming: scenario planning based on expert evaluation sets, benchmarks, and safety testing
- Multimodal and Multilingual: audio, video, and image data, plus collection in 150+ languages and dialects
- Off-The-Shelf Datasets: expert-authored data licensed for immediate training
“The response from the market for our services has been remarkable,” said Cory Elich, Executive Vice President of Operations at Guidepoint Labs, formerly with Invisible Technologies. “Within my first week, I was already engaging with several of the world’s leading frontier AI labs and their demand for proven and specialized expertise is even greater than we expected. We’re also seeing a growing appetite among our existing enterprise and institutional clients for AI training and evaluation. We’re building out an eight-figure pipeline quite aggressively and that tells us this need is real, urgent, and not being met by other players today.”
In its recently published Clinical Trial Design Benchmark, Guidepoint Labs assessed eight frontier models across 36 expert-designed tasks and 468 blind evaluations with 85+ credentialed Advisors averaging 18+ years’ experience. These experts found that only 5% of the tasks they evaluated from the models met their specialist standards. This benchmark highlights that while leading models can write trial designs that appear convincing at first glance, there are critical holes that can only be filled with true expertise. Guidepoint Labs aims to bridge that gap across critical domains and continue pushing the frontier forward. Explore the Clinical Trial Design Benchmark here.
Guidepoint Labs works with frontier model labs and applied AI companies building for specialized and regulated domains. Learn more at labs.guidepoint.com.
About Guidepoint
Guidepoint provides real-time access to expert insights, combining human expertise with AI-powered research tools to deliver knowledge at scale. Backed by a global network of more than 2M+ subject-matter experts, Guidepoint equips institutional investors, consulting firms, and corporations with the context they need across companies, markets, and trends. Through live, asynchronous, and agentic workflows, Guidepoint embeds expert knowledge directly into decision-making, turning answers into action when timing matters most. For more information, visit www.guidepoint.com.
About Guidepoint Labs
Guidepoint Labs provides expert human data for AI training and evaluation. Frontier labs and applied AI companies rely on Guidepoint Labs for expert-authored, validated data from credentialed professionals to build better models and agents across high-stakes domains, modalities and use cases.
Guidepoint Labs is powered by Guidepoint, a global pioneer in expert insights with a network of 2M+ vetted Advisors across 300+ industries and 150 countries. For more than 20 years, Guidepoint has delivered high-caliber experts, rigorous compliance and white-glove service to the most discerning institutional clients. Guidepoint Labs brings that same standard to AI teams. Please visit labs.guidepoint.com.
Contact
Meir Kahtan
Meir Kahtan Public Relations, LLC
[email protected]