In 2015 Renen Hallak walked out of a job he was good at and had nothing to show for it. He had been the first engineer at XtremIO, and later its VP of R&D, leading a group of several hundred engineers. The company had been sold to EMC, reportedly for about $430 million. He left anyway, with no product and no company, only a plan to ask questions.
For months he talked to storage buyers and storage makers about what they were stuck with. The same complaint kept coming back. Flash was too small and too costly to hold everything, disk was too slow, so everyone lived with tiers: a fast pile, a slow pile, and people paid to shuffle data between them.
A storage architecture cannot be shipped half finished, so VAST spent years in stealth building something that would either work or not.
What almost nobody notices is where he started. Hallak did not come to storage from hardware. He began his career in computational complexity research, a field nearer to mathematics than to disk drives, and joined XtremIO at its start in 2009. That habit of looking for a clean structure under a messy system shaped his bet.
The bet was specific. Cheaper QLC flash, better data reduction, a new kind of persistent memory for writes and metadata, and NVMe over fabrics might together let one flash system replace both tiers. In January 2016 he founded VAST Data with Shachar Fienblit, formerly CTO of Kaminario, and Jeff Denworth, who came from CTERA. Research and development stayed in Israel. Headquarters went to New York, close to the likely buyers.
Then came the long, quiet cost. A storage architecture cannot be shipped half finished, so VAST spent years in stealth building something that would either work or not. A $40 million Series A arrived in 2017 and a second $40 million round followed. When the company surfaced in 2019 it had hired the analyst Howard Marks as evangelist, which the trade press noted was his first full-time job since 1987.
Nobody could schedule what came next. After ChatGPT launched in late 2022, every large AI data centre needed to feed thousands of GPUs at once, and a design with no tiers looked like foresight. VAST raised $118 million at about $9 billion in 2023 and signed a $1.17 billion agreement with CoreWeave in November 2025. In April 2026 it raised about $1 billion at a $30 billion valuation.
Hallak rarely tells personal stories in public, and where he was born or studied is not in the record. What is there is a habit a first-time founder can copy: he asked customers what hurt before he wrote a line of company code, then waited years for the world to catch up.