Sign In

Shay Zweig

Tech leader | AI Consultant | Executive

Shay Zweig is an accomplished artificial intelligence (AI) and neuroscience professional, with a Ph.D. and M.Sc. from Bar-Ilan University. With an educational background in computer science, neuroscience, and machine learning, Shay is well-versed in the intersections of these fields. He has worked as a teacher assistant and lecturer at Bar Ilan University and The College of Management Academic Studies, respectively. He has also held multiple software development team leader positions with the Israel Defense Forces (IDF) and Jivy Group.

However, Shay is currently best known as the Chief Technology Officer (CTO) at Intuition Robotics, a company that he joined as the Head of AI. Here, he makes great use of his considerable expertise in AI and the application of machine learning to create unique, enjoyable, and personable digital experiences for seniors. In his role as CTO, Shay has played a crucial part in developing the company's award-winning social companion technology, ElliQ.

Highlights

Apr 26 · linkedin

Warning - a bit of a long post but here is the tl;dr: I built a really handy workshop titled "LLM for developers" that aims to introduce developers and technical product managers to the benefits of large language models and show them how to integrate this powerful tool into their skillset. This hands-on workshop uses code examples to explain the risks and benefits of using LLMs and introduces some of the tools, libraries, methods, and best practices for working with them. The feedback to the workshop was really positive and made me think that any R&D organization can benefit from it, especially if it’s accompanied by an LLM innovation hackathon.     If you think you think your organization can benefit from something of that sort, let me know.

Now the longer version (with some of my thoughts): One of the most interesting aspects of the recent explosion in GenAI is that it lowered the bar for innovating with AI. For years, we, the AI/ML/DS folks, enjoyed the fact that we have knowledge and experience (which we gained by lots of hard work) not many people possess, leaving the work of enabling AI driven products in the hands of few. The recent progress in our field somewhat changed this reality. Many of the models released (especially the better ones) are not open, and those that are, are big, complex and harder to train to a satisfying level without the resources of giant companies. However, many of the models are exposed in a very accessible API that allows a lot of flexibility.  So on one hand, great access to the best AI capabilities, and on the other hand, very limited access to the model internals to perform "classic" ml work, The outcome: This is maybe the best time in the past few years for non AI developers to contribute and create AI based products.   But, there is also a big catch here... One of ChatGPT's best qualities is its amazing interface which is responsible for much of its phenomenal success (and this is said by Sam Altman, OpenAI's CEO), however this is a double edge sword since it creates the illusion that you can achieve anything with this tool, and that the bar is not only lowered - it just vanished...  However, anyone who has attempted to create a product using LLMs knows that this is far from the truth. There are numerous concepts and practces one must understand to use this technology effectively in a real-world product and not end up with a mere gimmick.   The good news though, is that by gaining good understanding of some basic methods and principles (which is not super hard) you can reach a point where you can easily differentiate between bad ideas that are likely to fail as a product and remarkable ideas that you should pursue and nourish.     R&D organizations can enjoy a wave of innovation from their developers and product managers resulting in the incorporation of many LLM-based features into their products, and there is nothing happier than some democratization of technology!

#llm #chatgpt #innovation

Jan 8 · linkedin

For the past few months I’ve been working as an AI consultant with several companies and I think now it’s time to make it official.

I am an experienced AI developer, manager and executive, with a track record of building data science strategy and teams, solving hard problems with ML (In domains such as NLP, Computer vision, recommendation systems, RL and others), incorporating data driven solutions into production and communicating it to company executives.

I can help you:

  • design and implement data driven solutions to your organizations’ strategic problems in fields such as: Computer vision, NLP, Recommendation systems, Reinforcement learning, Generative models , Tabular and Time series data 
  • Build your AI strategy
  • Build and escort your internal DS teams
  • Communicate with the rest of the organization (including MLOps and other data functions such as analytics and data engineers)

If you are in need of such services, I’ll be more than happy to chat.

Related Questions

What specific projects has Shay Zweig worked on as an AI consultant?
How did Shay Zweig transition into AI consulting?
What industries does Shay Zweig specialize in with AI consulting?
What are some notable achievements of Shay Zweig in the AI field?
How does Shay Zweig integrate AI into organizational structures?
Shay Zweig
Shay Zweig, photo 1
Shay Zweig, photo 2
Add to my network

Location

Israel