Eli Singer works in the part of technology that almost nobody photographs. There are no consumer apps with his name on them, no viral launches. What he builds sits underneath: the machinery that turns a mountain of stored data into an answer a human can actually use. For more than twenty years that has been his single, consistent obsession, and it is the reason his name keeps appearing next to the hardest problems in enterprise data.
Today Singer is best known as a co-founder of Jethro, formerly JethroData, an Israeli-born company built around an idea that sounded almost reckless when it started: index everything. Where the rest of the big-data world assumed that indexes were too expensive to maintain at scale, Jethro fully indexed select datasets on Hadoop's file system so that a query would touch only the rows it needed, instead of grinding through a full scan of billions of records. The payoff was speed - the kind of sub-second response that makes an analyst forget there is a distributed cluster underneath the dashboard at all.
What he is building now
Jethro's pitch matured into something a business could say in one breath: a BI-on-Hadoop acceleration layer that speeds up query performance for the tools companies already use, including Tableau, Qlik and MicroStrategy. The promise was to analyze tens of billions of rows with sub-second response time while keeping existing infrastructure and front-end systems in place. No rip and replace. No new analytics team. Just the same questions, answered faster.
That restraint is the through-line of Singer's approach. Many competitors of the era - Cloudera's Impala, Hadapt, and the established analytics-database vendors - were fighting over whether to replace one platform with another. Jethro chose a narrower and harder path: leave the data where it lives and make it fast. It is an engineer's kind of ambition, measured in milliseconds rather than headlines.
Jethro is a BI-on-Hadoop acceleration layer that speeds up big data query performance for tools like Tableau, Qlik and MicroStrategy.
Jethro's founding value proposition
A career that never left the data
Singer's story does not begin with a startup. It begins at a terminal. Early in his career he worked as a database administrator at Morgan Stanley, the kind of role where you learn, in your hands, how large systems behave when they are pushed. That grounding shows up in everything he has done since. He does not treat data as an abstraction to be marketed; he treats it as a physical thing with weight and friction.
From there he became a founder. He started Memco Software and led it to a successful IPO on the NASDAQ, an early proof that he could take a technical idea all the way to public markets. He went on to co-found and lead WebCollage, a cloud company focused on e-commerce content that businesses used to syndicate and manage product information; it was later acquired by Answers.com. Along the way he held senior roles at Qualitas Health and served as chief executive of Zizio, moving fluidly between data, commerce and health technology without ever drifting far from the core theme of making information usable.
He studied management at Pace University, earning a bachelor of science between 1990 and 1993. But the more telling credential is the pattern: found something, build it, hand it forward, start again. Serial founders are rare. Serial founders who stay inside one hard, unglamorous domain for two decades are rarer still.
The Jethro bet
When JethroData raised its first $4.5 million round, led by Pitango Venture Capital, the framing was blunt. Companies were drowning in Hadoop data but starving for answers, because analysis ran in slow batches rather than in real time. Singer's team argued that the problem was not storage - Hadoop was fine at storage - but retrieval. Their fix was to process and index data natively so that queries never had to read more than they needed.
The company grew the way early startups do, planning to expand from a handful of employees toward two dozen as it moved from a single alpha customer to a private beta of its index-based SQL engine. In 2015 it raised an $8.1 million Series B led by Square Peg Capital, with Pitango returning, and adopted the shorter name Jethro. The founding team around Singer was deep: Boaz Raufman, the co-founder and CTO who architected the core technology after running big-data initiatives at Amdocs; Ronen Ovadya, a co-founder focused on customer success; and industry veteran Mark Kremer, who stepped in as chief executive as the company scaled, with Singer serving as President.
The goal is to analyze tens of billions of rows with sub-second response time, while keeping your existing infrastructure and front-end systems.
The Jethro performance promise
The quirk that made it work
There is a small, revealing detail in the Jethro design. The conventional wisdom in big data was that indexing every column was wasteful - too much storage, too much maintenance. Singer's team did it anyway, and turned the supposed weakness into the entire advantage. Auto-cubes and automated performance tuning, ideas that later got marketed across the industry as self-driving analytics, were part of the design early. It is a good window into how Singer thinks: take the thing everyone treats as too expensive to bother with, and check whether the trade-off actually holds at the scale that matters.
That instinct - to question a default rather than inherit it - is what carries across his companies. WebCollage made messy product data usable for retailers. Jethro made big data usable for analysts. Different decades, different stacks, the same job.
Watch: Jethro, explained
Singer has spoken at Hadoop industry events about why business intelligence struggles on big data and how an acceleration layer changes the equation.
The measure of the work
The best data infrastructure is invisible. When Jethro worked, the analyst never thought about Hadoop, never thought about indexes, never thought about the cluster. They asked a question and got an answer. That invisibility is Singer's real product, and it is a hard thing to sell precisely because success looks like nothing happening. His career is a long argument that the unglamorous layer - the plumbing beneath the dashboard - is where the real leverage lives.
He has done the full arc that the technology industry likes to celebrate: the IPO, the acquisition, the venture rounds. But the exits were never the point. The point was a stubborn, engineer's conviction, held across four companies and twenty-plus years, that data should move at the speed of a question. Everything he has built is a variation on that single sentence.
He started his data career as a DBA at Morgan Stanley before founding a single company.
His first company, Memco Software, traded on the NASDAQ under the ticker MEMCF.
Jethro's contrarian bet was to index every column - when the industry assumed that was too costly at scale.
He has founded or co-founded companies across data, e-commerce and health technology.