Iris.ai and AWS Enter Strategic Collaboration

Over the past decade, we have been building technology designed not for headlines, but for real-world complexity. Our mission has always been rooted in transforming fragmented, unstructured, large scale knowledge into reliable AI infrastructure.

We have built that technology, deploying it across 150 million documents for Fortune 500 manufacturers, European telecoms, and government agencies – long before most of this became mainstream. Now, the industry is starting to catch up.

We are proud to announce that Iris.ai has entered a strategic collaboration with AWS. This is an invitation-only designation from Amazon Web Services, reserved for companies demonstrating the most technically mature and commercially validated AI deployments on the AWS platform.

Why now, and why this matters

The enterprise AI landscape is crowded, but real, yet production-grade adoption remains rare. Most of what we see still doesn’t make it past demos, pilots or PoCs. The problem the way we’ve experienced is never lack of data - it is lack of context. Enterprises simply can’t get reliable results from AI systems when the underlying knowledge is fragmented, unstructured, lacking context, or locked in formats those systems cannot read.

As our Co-Founder, Anita Schjøll Abildgaard, puts it:

"We started in 2015 because the problem was already real: patents and documents no one could search, R&D insights locked in formats no system could read... We spent a decade building systems that worked on messy real-world data, deploying in steel manufacturing, pharmaceutical R&D, and government agencies where getting it wrong had real consequences. The AWS recognition is not a signal that we caught a wave. It is confirmation that we were right about the problem before the wave even existed."

This partnership positions Iris.ai as a preferred partner for AWS's data & AI modernization initiatives globally. It means we are tackling the exact problem that stalls enterprise cloud migrations: moving the data that matters most, scanned engineering specs, contracts, patents, and compliance filings, which cannot be handled by conventional ETL tools.

What this means technically and commercially

The AWS recognition is not ceremonial; it comes with direct commercial implications and deep technical integration. And this isn’t just a label – it changes how we operate, both technically and commercially:

• Native Integration: Iris.ai's platform runs natively on Amazon OpenSearch, Amazon Bedrock, EC2 GPU, and Sagemaker instances.
• Reduced Procurement Friction: We are now a Marketplace-enabled partner. Enterprise deals can be directly applied toward existing AWS Enterprise Discount Program (EDP) commitments, removing a common barrier for large-scale deployments.
• Global Expansion: We are serving as a Global Launch Partner in the Kingdom of Saudi Arabia marketplace region, alongside Dublin and US East and West.

This collaboration represents a fundamental shift in how enterprise AI is deployed. As Matthew Thomson, Director of EMEA Startups at AWS, noted:

"Our collaboration with Iris.ai brings together their powerful knowledge extraction and contextualization capabilities with the security and agility of AWS. This means that enterprises can put their most valuable data to work with high-quality LLM outputs, faster migrations, sharper decisions, and AI that runs in production and drives real outcomes."

Where we go from here

We have never been interested in surface-level transformation. Most enterprises run strong models on weak knowledge and wonder why the results disappoint.

AWS powers the compute and the models, while Iris.ai powers the knowledge those models reason on. By focusing on getting enterprise knowledge clean, contextualized, structured, and current before a model is ever invoked, we ensure that what those models reason on is actually reliable.

This AWS recognition isn’t just a milestone, it’s a signal. We are here to scale with intention, partner more deeply, and continue delivering the kind of AI that doesn’t just look good in demos, but performs where it matters most.

Let's make AI deliver.

➡️ Download the official press release (PDF) for full details.