JETHRO / JETHRODATA. The teal wordmark of an Israeli big-data startup that spent its life on a single, unglamorous problem - making enterprise dashboards answer in seconds. Herzliya & San Francisco.
Big Data · Business Intelligence
The engine that made interactive business intelligence run fast on big data - by doing the one thing everyone else had given up on: indexing.
Everyone loved big data until they ran a dashboard on it and started waiting. Jethro looked at that pause and asked a boring, brilliant question.
Hadoop made a simple promise: store everything, cheaply. Enterprises took the deal and poured billions of rows into their clusters. Then they pointed a business-intelligence tool at all that data, clicked a filter, and watched the spinner turn. The storage was cheap. The answers were slow. That gap - between having the data and being able to ask it a question - is the problem Jethro was built to close.
Founded in Israel around 2012 by Eli Singer, Ronen Ovadya and Boaz Raufman, JethroData started from a contrarian observation. Most of the SQL-on-Hadoop engines of the era - Impala, Hive, Presto, Drill - answered queries by scanning, reading through enormous amounts of data to find the rows that mattered. Jethro's founders, veterans of Memco, WebCollage, Amdocs and IDF Intelligence, reached back to an older database idea the field had largely abandoned: the index. Index every column automatically, and a query reads only the rows it needs instead of the whole dataset.
The result was an engine the company said could run queries up to 100 times faster than scan-based alternatives. But speed alone was not the pitch. Jethro's real argument was about who gets to use big data. "Interactive BI should be available to all business decision makers," Singer said - not just data scientists, not just the engineers who could hand-tune a cluster, but anyone with a question and a dashboard.
To make that true, Jethro chose to be invisible. Rather than replace the tools people already trusted, it sat underneath them. Tableau, Qlik and MicroStrategy connect to Jethro over standard ODBC/JDBC, with no changes to the application and no re-engineering of the data. A business user keeps their familiar dashboard; Jethro quietly makes it fast.
Investors bought the focus. Pitango Venture Capital led a $4.5 million Series A in 2013; Square Peg Capital led an $8.1 million Series B in 2015, bringing the total to $12.6 million. The money went into sales, marketing and deepening the engine rather than sprawling into new product lines - a company content to solve one hard thing well.
By 2016 the engine had grown three complementary tricks - full indexing, Auto-Cubes that automatically materialize common aggregations, and a result cache - so that dashboards increasingly hit small pre-built structures instead of raw data. The proof showed up in production: Tata Communications built an operational-intelligence dashboard on Jethro to track content-delivery performance across roughly two billion service requests a day, without a team of engineers hand-maintaining cubes. Avis, RBC, Symphony Health, iBasis, BICS and Fiat Chrysler are among the enterprises that put the engine to work across telecom, finance, healthcare analytics and automotive.
Jethro doesn't ask you to move your data or swap your BI tool. It slots in the middle and does the hard part.
Tableau, Qlik, MicroStrategy talk SQL over ODBC/JDBC.
Full indexing, Auto-Cubes and result cache turn queries into seconds.
Data stays put in HDFS or Amazon S3 - no re-engineering.
Every column is indexed automatically, so a query reads only the relevant rows instead of scanning the entire dataset. This is the core of the up-to-100x speedup.
Jethro detects repeated aggregation patterns and materializes them into small cubes - killing the manual cube-building that used to fall on data engineers.
Repeated and similar queries return instantly from cache, keeping dashboards responsive even under thousands of concurrent users.
Interactive BI should be available to all business decision makers- Eli Singer, Co-Founder & CEO
TechCrunch once called it Hadoop's "Achilles heel": the data was cheap to store but slow and batchy to query. Traditional fixes meant either buying an expensive data warehouse or paying engineers to hand-build cubes and pipelines. Neither scaled to thousands of business users asking ad-hoc questions.
Where rivals scan, Jethro indexes - a search-engine approach applied to enterprise analytics. It stays data-source agnostic (HDFS or S3), plugs into existing BI tools unchanged, and automates the cube and cache work that used to be a person's job.
Serial founder - previously co-founded Memco Software (NASDAQ) and WebCollage.
15+ years running large customer projects, including at Amdocs.
Big-data performance expert from Amdocs and IDF Intelligence.
Large organizations with big datasets and many concurrent BI users.
B2B enterprise software. Jethro licenses its SQL acceleration engine to organizations running BI on Hadoop or cloud object storage, distributed partly through channels like the AWS Marketplace and BI-vendor partnerships. Its value is tied directly to performance and cost - interactive analytics without a separate high-cost warehouse or hand-built cubes.
Jethro competed with scan-based SQL-on-Hadoop engines (Impala, Presto/Trino, Drill, Spark SQL, Hive), OLAP-acceleration layers (Druid, AtScale, Kyvos) and, at the high end, traditional data warehouses (Teradata, Vertica) that it positioned itself as a cheaper alternative to. Its edge: a search-engine indexing approach and a refusal to make users migrate.
Singer, Ovadya and Raufman set out to fix slow analytics on Hadoop.
Pitango leads; the team grows from ~8 toward 25, with one alpha customer and a beta on the way.
Jethro opens a private beta of its index-based SQL engine for Hadoop.
The BI-acceleration solution launches broadly; Square Peg Capital leads the round.
The engine adds automatic cube materialization on top of indexing and caching.
Jethro pairs its engine with additional enterprise BI analytics.
See Jethro sit underneath Tableau over an ODBC connection and accelerate dashboards on Hadoop.
Watch on YouTube ▶Browse product walkthroughs and BI-on-big-data explainers from the Jethro team.
Browse videos ▶Profile compiled from public sources including Jethro/JethroData press releases, TechCrunch, insideBIGDATA, BigDATAwire, The Silicon Review and Crunchbase. Figures such as revenue and employee count are approximate. Performance claims ("up to 100x") are the company's own published statements.