Thira raises $21M for self-running back offices
Former Apptio leaders are building governed AI agents that coordinate complex enterprise processes across existing systems.
Thira is entering the enterprise-agent market with an unusually large seed round and a thesis that companies need more than isolated AI assistants: they need systems capable of running entire back-office processes.
What happened
Thira raised a $21 million seed round led by Madrona, with FUSE and several enterprise executives participating. The company was founded by former Apptio leaders Sunny Gupta and Kurt Shintaffer, alongside executives and engineers from Atlassian, Oracle and Databricks.
Its initial focus is enterprise IT. The platform connects existing systems and uses governed, self-learning agents to execute workflows involving multiple steps, decisions and approval rules.
Why it matters
Many enterprise agents are effective at generating text or completing one task, but humans still coordinate the broader process. Thira is attempting to automate that coordination layer while preserving controls around permissions, oversight and business logic.
The size of the seed round reflects both the founders’ track record and investor interest in platforms that can move from AI assistance to operational execution.
The bigger picture
Enterprise AI is shifting from copilots toward systems of action. The long-term opportunity is not simply helping an employee work faster, but allowing governed software agents to manage processes across fragmented enterprise tools. Success will depend less on model novelty and more on integration, reliability and accountability.