Exited Unomaly founders raise $3M to launch Opper AI

The team behind Unomaly, the machine learning-based observability platform acquired by LogicMonitor in 2020, has re-emerged with a new mission: to make AI infrastructure as reliable as Stripe is for payments.

Opper AI announced the launch of its Task Completion API, backed by an oversubscribed $3 million pre-seed round. The API is designed to help developers move beyond one-off prompts and fragile prototypes, providing a structured and dependable path to launching AI-powered features in production environments.

The funding round includes backing from Luminar Ventures, Emblem Venture Capital (backers of one of the fastest growing AI startups in Europe – Loveable), Greens Capital, and several prominent angel investors. The Opper founding team – Göran Sandahl, Johan Gustafsson, and engineers from the Unomaly core team – bring more than 50 combined years of experience building automated, production-grade systems.

“The success of chatbots, agents, or data tasks comes down to how reliably they interact with models,” said Göran Sandahl, Co-Founder and CEO. “Opper removes the guesswork from prompt engineering and gives developers a tool that just gets the job done.”

A new infrastructure layer for AI developers

The Opper API introduces a new abstraction: task contracts. Instead of writing and tweaking prompts manually, developers define tasks in JSON; specifying inputs, expected outputs, and success criteria. The Opper API takes over from there, using smart prompt generation, built-in feedback loops, and model interoperability to deliver accurate and repeatable results.

Key capabilities include:

  • Write tasks, not prompts:

Developers define tasks in structured JSON and let Opper handle the prompting behind the scenes.

  • Built-in evals and metrics:

Quality and reliability improve automatically through integrated evaluation and feedback loops.

  • One endpoint, 80+ models:

Access to leading open-source and frontier models via a single API, with smart fallback routing to ensure success.

  • Full observability:

Track every token, cost, and model interaction in detail to optimize workflows and control spend.

Early traction and real-world use

Early adopters include companies like GetTested.io, which used Opper to streamline their AI pipeline for automated blood test reports:

“We traded a patchwork of home-grown prompts and frameworks for Opper,” said Adam, AI Engineer at GetTested. “It clearly works – and we’re now rolling out our pipelines across 62 countries including the US and India.”

“We’re seeing again and again that our customers can go confidently from beta to reliable production deployments,” added Johan Gustafsson, Co-Founder and CRO.

Availability and pricing

Opper AI’s Task Completion API is available now. Developers can get started with $5 in free inference per month, with paid plans starting at $5/month and enterprise pricing available on request.

For more startup news, check out the other articles on the website, and subscribe to the magazine for free. Listen to The Cereal Entrepreneur podcast for more interviews with entrepreneurs and big-hitters in the startup ecosystem.