Who’s watching the algorithms? 3 Israeli startups predicting performance of AI models

Three Israeli startups tackling the blind spots of AI dependency – real time performance monitoring, reducing bias and AI explainability.

It’s the opposite of an “AI winter” it seems, as investment levels in Artificial Intelligence continue to break records globally. According to the State of European Tech 2019 report by Atomico, European AI companies raised $4.9 billion in 2019 alone.

State of Euorpean Tech 2019, AI investments

When it comes to AI however, it’s mostly a black box. Data scientist know that the accuracy of the models to make predictions or recommendations, has to do with data quality/annotation and training of the algorithms, but most simply trust the data and thus create a blind spot. As more companies embrace AI for their needs, what happens when the algorithms start slipping? who’s policing the quality?

The stakes are high for AI model performance, as it may affect decisions taken on loans, patient care, cyber security, autonomous vehicles and so on. You can imagine the scenarios where a deterioration in the confidence level of the models can create a real mess.

Several Israeli startups understood there is a potential market in monitoring AI performance and are tackling this market. While these solutions don’t provide 100% “explainable AI” that opens the black box, they solve the problem of 3rd party verification of quality.

Superwise.ai – started in January 2019, Superwise hopes to become the AI Performance Management for the enterprise, connecting to any ML system in production. It offers real-time performance monitoring & analysis, predictive intelligence that is based on machine learning models, auto-benchmarking, out of the box explanation and bias detection and cloud or on-premise configuration. Superwise raised an undisclosed round from F2 Capital, and is founded by two brothers, Oren and Ofer Razon. Ofer the CEO previously started the AI unit at Amdocs.

Superwise dashboards – source

MonaLabs – founded in October 2018, Mona labs is targeting data science teams. It aims to maintain data integrity to solve the garbage in, garbage out risk when data is mis-formatted. Mona also tackles model bias and detects concept drift. It offers a performance benchmarking score and allows to evaluate the quality of models in production. It works in the cloud or on-premise as well. Mona announced an undisclosed seed round led by Global Founders Capital in Feb 2019. Started by a team of ex-Googlers, who previously led the work on Google Trends.

MonaLabs dashboard – source

Deepchecks.com – focus on warning organisations before their deployed AI systems make bad calls using proprietary predictive analytics algorithms. The company was backed by Grove Ventures in an undisclosed round (still in stealth, so not much information available yet).

deepchecks.com
Deepchecks, still in stealth (source)

For a deeper technical review of other services for data scientists, check out this excellent Medium post by Ori Cohen.

Follow me
Co Founder and Managing Partner at Remagine Ventures
Eze is managing partner of Remagine Ventures, a seed fund investing in ambitious founders at the intersection of tech, entertainment, gaming and commerce with a spotlight on Israel.

I'm a former general partner at google ventures, head of Google for Entrepreneurs in Europe and founding head of Campus London, Google's first physical hub for startups.

I'm also the founder of Techbikers, a non-profit bringing together the startup ecosystem on cycling challenges in support of Room to Read. Since inception in 2012 we've built 11 schools and 50 libraries in the developing world.
Eze Vidra
Follow me
Total
0
Shares
2 comments
  1. Great writing and timely opinion. But as you know everything has a cycle. Even the car after more than 100 years seem to be at the end of it’s cycle. Will we see the day of the AI phase goes down? of course we will if it’s five to ten years away, that’s fine. Just as mobile phones have changed almost everything in communication, so will AI change almost everything that needs intelligence. Not just cars and homes, but also communication and definitely “content”. Great to see again Israelis hard at work in a new technology phase… it will again change how the country looks and grows. Thanks again for an interesting read! #amivider

Comments are closed.

Previous Article

The stars align for AR to go mainstream

Next Article
Israeli high tech talent in high demand

Israel's talent crunch detailed in new report

Related Posts
Total
0
Share