Bay Area, California


News-Driven Quantitative Trading


Prediction Machine is a team of data engineers, DevOps, data scientists, and successful investors. We are a new company formed by experienced founders. Rather than imitate, we prefer to pick less explored problems and work from first principles. Our goal is to ship investment strategies that outperform because they leverage unusual data, or usual data in an unusual way, and rely on massive computational abilities to make plausible, sample-efficient estimates.


We believe algorithmic portfolio management will undergo a dramatic change in the coming years. We are here to be a part of this change. Every reinvention begins with a blank sheet of paper, and a willingness to reason from first principles.
This is where Prediction Machine enters the picture. Our first move in this direction was a global search for talent. To paraphrase Elon Musk, we are building the team that will build the solution.

Selectivity and Synthesis

We are searching the world for the right mix of talent, drive, and risk-taking. We are looking for people who have prepared to be lucky.


We’re not nine-to-fivers. We’re looking not for employees, but for those who can ‘ship it’ and also be intellectual partners — data scientists who share our curiosity and quest to uncover the underlying truths; data engineers who aspire for their code to be like poetry.

Our story

How It All Started

Tim Chklovski, our founder, has an extensive background in AI, mathematical problem solving, coding, and data science. After working at startups and larger companies, he went all-in on tackling stock markets in 2014 in his first fund, which focuses on developing high-conviction views about companies likely to outperform for the long term.
Over time, he extended his focus to how various developments and stock-related news impact market prices. Prediction Machine represents doing this analysis justice — discovering quantitative opportunities in the markets requires a team of data scientists, data engineers and devops to harness large compute to efficiently formulate and test hypotheses about the market.


At Prediction Machine, we recognize that to solve a problem well, we must first see it well, and to see something others are not seeing. Our machine learning techniques must spot patterns and connections before they become clear to other algorithms or humans. That requires doing things differently, not being afraid to try new ideas that are “out there.”


We’re not just in it for the money. We want to develop strategies that deliver better profits, but over the long term, we aim to develop tools and techniques which bring computers to analyze and develop insights about the world we live in. We aim to give back in the fields of computer interpretation of information, model induction, as well as in tools that help harness and direct certain types of large computations. We think in the future, most developers will be defining search spaces for computers to explore, rather than hand guide useful paths in these spaces.

Our Team

Timothy Chklovski

Founder and Managing Partner

Tim is a medallist of the International Math Olympiad, having represented the US in the 1990s. He holds a PhD from MIT in Computer Science (Artificial Intelligence) and an SB in Math, as well as an SB and an MEng in EECS (all from MIT).
After founding an Accel-funded web analysis company in 2000, Tim was a research scientist in the 2000s. He focused on crowdsourcing and web extraction techniques, then became Founding Engineer and Chief Scientist at Factual, which merged with Foursquare, and later worked on large-scale ML in the Revenue team at Twitter.
Tim made his life-long passion for stock markets ‘official’ by launching his first fund, Prepared Mind Advisers, in 2014. Seven years later, Prediction Machine, founded in 2021, is the result of his doubling-down on the opportunities that he believes large compute and alternative data will offer in uncovering pockets of predictability in markets.
Having immigrated from Russia in his teens, Tim has lived in the Midwest, Boston, Los Angeles, and for the past decade in the Bay Area. He enjoys self-improving algorithms and elegant code.

Charles Stephen Wallman

Strategy and Culture Partner

Steve launched his investment fund in 1994, outperforming the markets by a wide range in the ensuing 26 years. He achieved what very few money managers achieve by out-thinking his competition (easier said than done!).
Steve enjoys ‘going Hilbert’, taking an unusual amount of effort to get to the very essence of things, usually by seeing them for what they are, based on first principles. Through extensive research, thought, and preparation, he was able to both identify early and hold such long-term successful investments as Apple, Nike, Intuit, and Interactive Brokers.
Steve also founded StudyBlue, an educational start-up where he grew and mentored its team. It was later acquired by Chegg, a connected-learning platform.
Being a contrarian at heart, Steve became one of the main investors and GPs of Prediction Machine in 2021. He focuses on thinking through key sources of Prediction Machine’s competitive advantage, and how its approach can differ. He also reminds us about the golden rule and focusing on bringing out the best in people, which we regard as our way of earning our right to win.
Steve lives away from the bustle in the Midwest..

Gil Elbaz

Data and Strategy Partner

Gil is an eminent tech entrepreneur. He and Tim go back nearly two decades, having worked together at Factual where Gil was founder and CEO, and Tim was one of two founding engineers.
After graduating from Caltech, Gil co-founded Applied Semantics. The company was acquired by Google in 2003, contributing to the AdSense product and forming Google Santa Monica. In 2007, he founded Los Angeles-based Factual Inc., an open data platform founded to maximize data accuracy, transparency, and availability.
In 2021, driven by his long-held interest in the markets, Gil has joined Prediction Machine as one of its three Founding Partners.
At Prediction Machine, Gil focuses on guiding the business through its growth trajectory, as well as scouting and emphasizing promising data opportunities.
Gil has served on the Board of Trustees for the XPrize Foundation and Caltech. He is also a member of the Los Angeles Social Venture Partners, and he is a partner at TenOneTen Ventures, a venture capital fund. He and his wife, Elyssa, also manage the Elbaz Family Foundation, supporting environmental and educational causes.


Join the team of top-notch professionals at Prediction Machine 

Get in Touch


Contact Us
Close Bitnami banner