Ocient has closed a $100 million Series C funding round to expand its agentic AI analytics business, the Chicago company announced on Thursday. The round was co-led by BOKA Capital and OCA Ventures, with all of the company’s existing investors taking part, and brings Ocient’s total capital raised to $260 million. Ocient says the money will accelerate delivery of its OcientAIQ Solutions, which let AI agents and human analysts query petabyte-scale datasets without sampling them first. The company expects its revenue to more than triple in its current fiscal year.
What the money is for
The funding lands as AI agents become the heaviest users of enterprise data. Ocient’s platform is built for organisations whose datasets are measured in petabytes and keep growing, including national security agencies and telecommunications providers that need to analyse network records. Its pitch is that sampling, the usual shortcut for huge datasets, misses the rare events that matter most.
Demand is coming from customers Ocient already serves. Chief executive John Morris said much of the growth comes from new customers in national security and telecommunications, particularly organisations consolidating cybersecurity information. Like Sentra, which raised $150 million to govern what agents can reach, Ocient sells control as much as speed: governance, lineage and audit trails built into the engine.
Agents are now the customers
The most telling detail is who is doing the querying. Morris said software agents, rather than human analysts, are increasingly the ones interrogating Ocient’s systems, driving higher utilisation and demand for more comprehensive datasets. That shift changes what a data platform must do: answers have to be reproducible and auditable, because no human watched the agent ask the question.
Morris was blunt about the problem his customers face. They have been accustomed to sampling data coming off the network, he said, because nothing they owned could handle the ingest rates or store tens of petabytes. Ocient runs analysis inside the engine rather than through a language model, so every result can be reproduced and audited with minimal data movement.
Our read
This round belongs to the same shift as the rest of the autumn’s infrastructure funding. Applied Compute raised $350 million to tailor models for specific businesses, and Ocient is the other side of that trade: the data layer those models and agents have to read.
Investors are no longer just backing the models; they are backing the plumbing that makes agents safe to run on real enterprise data. The bet is that governance becomes the selling point, not the constraint, once agents start touching regulated records at machine speed.
The numbers deserve a sceptical reading. Ocient expects revenue to more than triple this year, after doubling the previous year, but it discloses no absolute revenue figure and no valuation. Morris said the funding should be enough to reach breakeven, but declined to give a timetable. Until the company publishes numbers, the tripling claim is a direction, not a destination.
Why It Matters
The agent era has a data problem that model releases alone cannot fix. Agents query constantly, at machine speed, and they inherit whatever access the humans before them held. A platform that lets agents read complete datasets, with governance enforced inside the engine, removes the sampling compromise that has limited large-scale analytics for years.
Nooks raised $500 million for AI sales agents in the same week, and the pattern is consistent: the money follows the agents, then the infrastructure that keeps them honest. For enterprises, the practical test is whether governed, full-resolution data lets agents act faster without creating a breach that only surfaces after the fact.
Background
Ocient, headquartered in Chicago, has now raised $260 million in total. Earlier this year it formalised its India operations and a dedicated national security subsidiary, and it plans to expand in Europe, Asia-Pacific and Latin America. General Compute raised $100 million in the same season for non-Nvidia inference chips, another bet that AI money is moving from models to infrastructure.
Source: Ocient’s funding announcement and SiliconANGLE’s report
0 Comments