Multithread effectively and personalize outreach to convert deals faster
Elevate social presence and drive business growth from social media
Identify and prioritize high-intent leads, and improve sales effectiveness
Find and connect with ICP attendees, and improve event outcomes
Nicky Penkov is a Staff Research Scientist at Google DeepMind, focusing on robotics and artificial intelligence. His research is centered on developing robots that can learn complex manipulation skills and generalize these skills across a wide variety of tasks, objects, and environments. He is particularly interested in imitation learning, reinforcement learning, and the creation of large-scale, general-purpose robotics models (often referred to as foundation models for robotics). Nicky aims to bridge the gap between the vast knowledge available on the internet and the physical capabilities of robots, enabling them to understand and interact with the world in more human-like ways. He completed his PhD in Robotics at the Georgia Institute of Technology and a postdoctoral fellowship at the University of California, Berkeley, before joining Google.
Nicky Penkov's work history includes a series of influential roles in various companies. Here is a detailed list of his professional journey:
Played a significant role in the development of RT-2, a novel vision-language-action (VLA) model at Google DeepMind. RT-2 co-fine-tunes large language models with robotic data, allowing it to transfer conceptual understanding and reasoning from web data to robot control, demonstrating emergent capabilities and improved generalization in robotics.
Led and contributed to the creation of RT-1, a multi-task, multi-embodiment transformer-based model for real-world robotic control. RT-1 demonstrated the ability to learn from large, diverse datasets of robot experiences and generalize to new tasks and situations effectively.
His research has consistently pushed the boundaries of how robots can learn and generalize skills, focusing on techniques that allow robots to perform a wide array of manipulation tasks without needing to be retrained for every new scenario. This includes work on data-driven approaches and scaling robot learning.
Earned a PhD with a dissertation focused on scalable and data-efficient deep reinforcement learning for robotic control, laying groundwork for his later contributions to large-scale robot learning.
CIPD Qualifications
Kingston University
University of Southampton
Putney High School
Highperformr Signals uncover buying intent and give you clear insights to target the right people at the right time — helping your sales, marketing, and GTM teams close more deals, faster.
Dotdigital is a customer engagement platform that empowers marketers to connect customer data, surface powerful insights, and automate intelligent messages across email, SMS, social media, and more. It helps businesses to acquire, convert, and retain customers through personalized and automated cross-channel communications, ultimately aiming to enhance customer loyalty and drive revenue.
Get verified emails, phone numbers, and LinkedIn profile details
Discover contacts with similar roles, seniority, or companies
Uncover insights like skills, work history, social links, and more
Explore contacts in-depth — from verified emails and phone numbers to LinkedIn activity, job changes, and more — all in one powerful view.
Highperformr AI helps you surface the right people and enrich your CRM with live, accurate contact insights so your teams can connect faster and close smarter.
Thousands of contacts — including decision-makers, influencers, and ICP matches — are just a search away.
Thousands of companies, including, are just a search away.