New technologies fold space. Possibilities that were distant become adjacent, and the maps we have no longer describe everything within reach. We explore those changes in Folding Space, describing opportunities where we think one or more companies could now be built.
We are investors who write six- and seven-figure checks ourselves and help founders build companies as well as raise additional capital. Our taste runs to the nerdy end of the market: we want to understand the technology, the potential market, what could make the business defensible, and the people building it. We bring decades of experience building and funding companies, and we like working with founders who teach us something new.
Who We Are
Harish and Anand are entrepreneurs, investors, and business leaders who have stayed close to their technical roots in computer science. We share a focus on technology venture investing, having arrived there through startups and academic research in Anand’s case, and financial markets and systematic trading in Harish’s.
Anand Rajaraman co-founded Junglee (acquired by Amazon, where it seeded the third-party marketplace) and Kosmix (acquired by Walmart, where it became WalmartLabs), and co-invented Amazon Mechanical Turk. He is a founding partner of Cambrian Ventures and Rocketship.vc and an early investor in Facebook and Lyft. He co-wrote Mining of Massive Datasets with Turing Award winner Jeffrey Ullman, and his research has received three retrospective ten-year best-paper awards. He teaches CS141 (Sports and Data) at Stanford and co-owns the San Francisco Unicorns cricket team.
Harish Devarajan is a technology investor whose institutional and personal investments span major AI platforms (OpenAI, Anthropic, Perplexity, Cohere), semiconductors (Cerebras, Tenstorrent, Lightmatter), deep tech (SpaceX, Commonwealth Fusion Systems) and other technology sectors (Apptronik, Reddit, StarkWare). He spent two decades on Wall Street, serving as a trader and COO on Deutsche Bank’s largest equities proprietary trading desks in New York and then leading Deep Value, a systematic trading firm that at times accounted for more than 5% of daily US stock-market volume. Harish holds advanced degrees in computer science and mathematics from the MIT and Stanford, and attends technical conferences for fun.
What this Blog is About
We write about technologies nearing commercialization. We also connect the dots between new technical capabilities and the large markets they could open. Founders working on these problems will see things we cannot. We publish our thinking to contribute to that exploration and learn from it.
When we write about startup ideas in detail it is not because we are wedded to the proposed technology approach and business model. We want to be clear enough to be wrong (and hope you will correct us before the market does!) and think big enough that we inspire the big thinker in you.
We write for people who could build the kind of companies we discuss. Mostly that means researchers and engineers already deep in AI, the kind who present or can read ICML and NeurIPS papers and wonder whether there is a startup there. It also means the operators and forward-deployed engineers who know how a bank or an insurer works, and who can get a pilot signed. The best teams we have seen combine technical depth with a close understanding of the problem.
We will sometimes write about possibilities enabled by technological changes that don’t map neatly to a startup idea. Often these might be in areas we are passionate about. Occasionally we will speculate about possible futures that are not currently within reach, but we think are super-cool anyway.
We write for people who could build the kind of companies we discuss. Mostly that means researchers and engineers already deep in AI, the kind who present or can read ICML and Neurips papers and wonder whether there is a startup there. It also means the operators and forward-deployed engineers who know how a bank or an insurer actually works, and who can get a pilot signed. The best teams we have seen combine technical depth with a close understanding of the problem.
The Name
In the Dune universe, folding space brings distant places within reach. That is what a good technology does to a market, and what we hope a good map does for a founder. We will indulge in the occasional science-fiction reference as we are unabashed fans of the genre, and we’ve invested in monumentally successful startups whose business plans could pass for sci-fi plot summaries.
Reach us
If you are exploring one of these opportunities, think we have missed something, or are working on a possibility we have not written about, we would like to hear from you.
Write to investors@foldingspace.ai or use this link. Tell us what you’re building or thinking of building, and share one or two of the coolest or most technically challenging things you or anyone on your founding team has ever done. They don’t have to relate to your startup. We’d especially welcome links to the work—GitHub repositories, research papers, Google Scholar profiles, faculty or academic webpages, technical demos, or engineering write-ups. You don’t need a finished product or a fully formed company to start the conversation.
We read everything. Some such first introductions have led to checks from us and great companies.