NitronEdge TechnologiesHelping Enterprises Harness Innovation
Data · Denodo

Denodo signs AWS deal to feed live data to AI agents

Denodo and AWS have signed a multi-year strategic collaboration agreement for agentic AI. What it covers, why it matters and what data teams should do.

Denodo announced on 10 November 2026 that it has signed a multi-year strategic collaboration agreement with Amazon Web Services. The deal is aimed at one problem: giving AI agents built on AWS reliable, live access to company data that sits in many places, without copying all of it into a new store first. It builds on an existing relationship and commits both companies to more joint work on the Denodo Platform on AWS around the world.

What changed

A strategic collaboration agreement is a commercial commitment, not a new product. Under this one, Denodo and AWS plan to run workshops that help customers pick agentic AI use cases and design them around their own data. They will also run joint marketing, align their field sales teams to sell together, and offer commercial incentives for technical evaluations and for purchases through AWS Marketplace.

The two companies also plan packaged offerings with systems integrators for specific agent-driven processes. Denodo names three examples: fraud prevention in banking, claims handling in insurance, and predictive maintenance in manufacturing.

On the technical side, Denodo says its platform lets AWS services such as Amazon Bedrock AgentCore, Amazon Quick and Amazon SageMaker reach enterprise data held on premises, in SaaS applications, in sovereign clouds and on other public clouds. The approach is zero-copy: agents query data where it lives through Denodo's logical data layer, instead of waiting for each dataset to be moved. Teams can reuse the same data products, business definitions and access rules for many agents, rather than building a new data path for every use case.

Denodo is available as a managed cloud service, called Agora, or can be deployed in a customer's own AWS account. It is an AWS ISV Accelerate partner and holds AWS competencies in data software and financial services. The release includes a customer comment from Bayer Crop Science in Latin America, which says it cut the time its teams spent searching for data by 80% and now runs customer service and tracking agents in production.

Why it matters

Most agent projects stall on data, not models. An agent that answers a claims query or flags a suspicious payment needs current figures from several systems, often including an old core platform on premises. Copying all of that into one cloud warehouse before the first agent goes live takes months and creates new copies that must be secured.

A logical data layer offers another route. The agent sends one query, the layer fetches what it needs from each source, and the same security rules apply whichever agent or person asks. Business terms such as "active customer" or "open claim" are defined once, so different agents do not give different answers.

There are trade-offs. Querying live sources puts load on them, and some legacy systems cope badly. Latency matters for agents that take several steps. Caching and careful query design still apply, and some data will still be better moved into a lakehouse for heavy analytics.

The commercial parts of the deal also count. Marketplace purchasing lets firms with committed AWS spend buy Denodo through that commitment, which can shorten procurement. Funded workshops lower the cost of a first proof of concept.

What this means for organisations in the Middle East, India, Africa and North America

Data residency is the main reason this matters in our markets. Banks, insurers and public bodies in the UAE, Saudi Arabia and Qatar often keep core data in-country, either on premises or in a local cloud, because of national data protection laws and sector rules from central banks. The same applies to Indian firms under the DPDP Act and RBI directions on data storage, and to South African firms under POPIA. These organisations want to use AWS AI services but cannot move every record into them.

A zero-copy layer lets the agent and model run in an AWS region while sensitive records stay where the rules require, with only the fields a query needs passing through and under the same access policies. AWS runs regions in the UAE, Bahrain, India, South Africa and Canada, which helps keep that traffic local. Each design still needs a legal and security review of what crosses borders, including prompts and logs.

For Canadian financial institutions, OSFI guideline E-23 expects clear records of what data a model used. A single place where data access is defined and logged makes that easier to show. In the United States, the NIST AI RMF points in the same direction.

What to do next

  1. Pick one agent use case, such as claims triage or fraud alerts, and map every data source it needs, including where each one is hosted.
  2. Mark which sources must stay on premises or in-country, and design the data flow around those limits before choosing tools.
  3. Define the shared business terms and access rules once, in the data layer, so every agent uses the same meaning and the same permissions.
  4. Load-test critical source systems before letting agents query them live, and plan caching where needed.
  5. Ask your AWS account team whether the new workshops and Marketplace incentives apply to your evaluation.

To learn more about the platform, visit our Denodo partner page. For help designing the data side of agent projects, see our data integration services.

Sources

  1. Denodo Signs Strategic Collaboration Agreement with AWS to Accelerate Trusted Data Access for Agentic AI (2026-11-10)
  2. Denodo Signs Strategic Collaboration Agreement with AWS (Macau Business) (2026-11-10)
Related: Denodo
Enquire

Talk to us about Denodo

Get a free consultation from our AI, data and security specialists. Honest advice, clear options and a reply within 6 hours.

Reply timeOur team replies within 6 hours

Request a callback

Free consultation. Our team replies within 6 hours.

Your details are safe. Our team replies within 6 hours. By sending, you agree to our privacy notice.

WhatsApp