> ## Content Index
> Fetch the complete content index at: https://www.therecursive.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Constructor Capital Backs Ora Computing with €3.5M to Build the Efficiency Layer of AI
- URL: https://www.therecursive.com/constructor-capital-backs-ora-computing-ai-model-compression/
- Published: 2026-06-24T18:05:00.000Z
- Updated: 2026-06-25T08:06:31.000Z
- Description: Deep tech investor Constructor Capital co-led the round alongside Greencode Ventures and XISTA Science Ventures to back a startup that compresses AI models by up to 80%, cutting inference costs for companies deploying AI at scale.
- Author: Yoana Stanislavova
- Tags: Deals, Constuctor Capital, Deep Tech, AI

Swiss deep tech investor [**Constructor Capital** ](https://constructor.capital/?ref=therecursive.com)has co-led a **€3.5 million seed round** in Ora Computing, a Vienna-based startup that compresses large AI models to make them cheaper and faster to run.

## The investors

Constructor Capital is a Switzerland-based early-stage VC investing in deep tech and science spinouts. The round also includes Helsinki-based Greencode Ventures, focused on applied AI across energy, industry and infrastructure, and Austria-based XISTA Science Ventures, which served as a foundational investor in Ora from the start.

> "The era of brute-force AI is hitting its physical limits: hyperscalers reopening nuclear reactors, frontier labs burning billions on compute, reasoning models multiplying inference costs every quarter. The only sustainable answer is to make frontier AI dramatically cheaper to run. That's what Ora Computing does – compresses frontier LLMs so they run anywhere: in the cloud, in your car, in your pocket," says **Valentino Jadrisko, Senior Associate at Constructor Capital.**

AI inference has become a significant and fast-growing cost for companies deploying AI at scale. For large-scale deployments, compute costs can reach tens of millions of euros per month, and the problem compounds as models continue to grow in size. For companies that want to run AI locally on devices like cars or industrial machines, the models are often simply too large to fit. 

[**Ora Computing** ](https://www.oracomputing.com/?ref=therecursive.com)is an **Austrian startup specializing in optimizing and compressing AI foundation models**. Their software reduces model size by up to 80%, making them run four times faster, with an accuracy loss of just 0–5%. 

The startup has already proven this in practice – a 70B-parameter model was compressed in a matter of hours for under $1,000, while the industry typically pays hundreds of thousands for the same result. Unl**ike competitors like Qualcomm and Intel**, whose tools are tied to their own hardware, Ora works across different hardware types and requires no changes to existing infrastructure. 

## From quantum physics to AI efficiency

**Stefan Sack and Raimel Medina**, two quantum computing researchers from the Serbyn group at the Institute of Science and Technology Austria (ISTA), founded Ora Computing to tackle what they see as a more urgent problem than quantum computing itself. The company emerged from stealth at the end of 2025\. Their background in quantum physics sets them apart from most AI founders, who typically come from Big Tech or machine learning research. 

> "We founded Ora Computing to challenge the assumption that massive scale is needed to reach useful intelligence"says **Stefan Sack, CEO & co-founder of Ora Computing.**

The funds will be used for three things: growing the team, expanding compression capabilities to larger frontier models, and launching a commercial product for cloud inference providers and companies that deploy AI at the edge.

With [AI investment trends](https://www.therecursive.com/cee-ai-investment-trends-2026-the-shift-to-applied-ai-and-roi/) pointing toward efficiency and cost reduction over raw compute power, Ora Computing is positioning itself at the center of that shift.

> “AI's energy appetite is growing faster than the world can build the infrastructure to feed it. One key approach is to make AI itself more efficient, and that is exactly what Ora does. Compressing models radically without sacrificing accuracy makes a tremendous difference to their customers,” says **Terhi Vapola, Founder and Managing Partner of Greencode Ventures.**