Ocient has spent a decade arguing that analytics should never sample data. That argument is suddenly worth a lot: the Chicago company closed a $100M Series C, taking its total raised to $260M.
Chief executive John Morris says national security and telecom customers are driving the growth. Those buyers were previously forced to sample network traffic because nothing could ingest at their rates or store tens of petabytes affordably.
Agents changed the buyer
Where human analysts once ran scheduled queries, software agents now interrogate the platform continuously. That raises utilisation, expands the working set and pushes customers toward keeping the full history rather than a slice of it.
Ocient says revenue will more than triple this fiscal year ending January, after doubling last year. The new money funds OcientAIQ, the product line that lets agents and people query petabyte-scale datasets, and the company expects it to be enough to reach breakeven, though Morris would not say when.
The architecture is deliberately unusual. Data stays put instead of being shuffled between tiers, and forward error correction replaces replication, which cuts storage overhead. It competes with Snowflake, Databricks and open-source stacks, though it often sits underneath them.
Sovereignty pulls deals
Most current opportunities are on-premises, Morris says, as customers repatriate for cost, data protection or sovereignty. Government and defence work is large enough that Ocient spun up a dedicated national security subsidiary.
Cybersecurity accounts for a rising share, with telecoms and security teams using the platform to scan complete historical network records for threats that sampling would miss. Morris is careful about the label, noting Ocient supplies the data layer rather than detection logic.
Hiring will lean toward sales and customer delivery rather than pure engineering, a sign the bottleneck has moved from code to deployment.
