fileAI signs AWS deal to bring outcome-driven AI agents to banks and insurers

• Who: fileAI, the enterprise intelligence company behind fileForge, and Amazon Web Services (AWS).
• What: A strategic collaboration agreement (SCA) to bring a new suite of outcome-driven AI agents to banking, financial services and insurance (BFSI) leaders, built to complete critical operational workflows, not just assist with them.
• When: Announced October 26, 2026, in a release datelined Singapore.
• Platform: fileForge, built on Amazon Bedrock and available through AWS Marketplace, serving industry leaders across Asia and the US.
• Tools: fileScout maps files ahead of extraction, optimising model routing and reducing token consumption; fileForensics surfaces actionable fraud signals directly into the processing pipeline.
• Scale: fileAI says it has processed over 1 billion files, and counts MS&AD, Toshiba, PwC, KPMG, Nippon Paint and Keppel among the enterprises trusting it.

fileAI announced a strategic collaboration agreement with Amazon Web Services on October 26, 2026, to bring a new suite of AI agents to banking, financial services and insurance leaders. The pitch is a deliberate break with the chatbot era: these agents are built to complete critical operational workflows, not just assist with them — and to be judged on the finished, auditable result.

Files in, finished work out

At the centre of the deal is fileForge, fileAI's data intelligence platform. It turns enterprise files into what the company calls trusted, governed intelligence for AI workflows, pairing repeatable skills with business context so complex processes can be carried through to completion. Under the hood, fileForge unifies data capture, preparation, governance and orchestration into auditable, SOP-driven workflows that transform unstructured and semi-structured data into reusable process intelligence. It is built on Amazon Bedrock and available through AWS Marketplace, giving industry leaders across Asia and the US direct access to the agents with AWS's security, scale and infrastructure behind them.

That framing matters because the failure mode fileAI is selling against is familiar to anyone who has run an enterprise AI pilot: a demo that works on clean sample documents and falls apart on the real file room. fileAI's answer is structure before intelligence — governance and validation wrapped around the model, so the output can be traced, checked and reused rather than merely admired.

Two tools do the quiet work

Two of the company's latest innovations underpin the pitch. fileScout maps files ahead of extraction, optimising model routing and drastically reducing token consumption — a direct attack on one of agentic AI's least glamorous costs, the tokens burned reading documents that turn out to be irrelevant. fileForensics surfaces actionable fraud signals directly into the processing pipeline. Together, the company says, they ensure only clean, relevant, fraud-checked data ever reaches the workflow. For AWS customers, that means faster cycle times, fewer exceptions and less manual rework, with the governance and traceability that regulated organisations require.

Why banks are the proving ground

Banking, financial services and insurance are an unforgiving first market for agentic AI: the workflows are document-heavy, the regulators want a paper trail, and a wrong answer is not a curiosity but a liability. That is why both companies framed the announcement around trust rather than capability. “Enterprise AI has to be judged on outcomes, not activity,” said Christian Schneider, CEO of fileAI. “By building on Amazon Bedrock and deepening our collaboration with AWS, we're giving finance and financial services leaders agents that take their most complex data and carry the work through to a completed, auditable result.”

Jamie Simon, Director of Partner Sales and Development for Asia Pacific and Japan at AWS, put the customer problem the same way: “Financial services customers consistently tell us that getting unstructured data ready for workflows they can trust is one of their biggest challenges.” fileAI, he said, combines document intelligence with the governance regulated enterprises require, and the collaboration makes it easier for customers to move from AI pilots to production with confidence — with AWS looking to scale it across the region.

The runway behind the deal

The agreement follows a busy few months for fileAI. On August 27, the company announced investment from SMBC Asia Rising Fund and Singtel Innov8 alongside the debut of fileScout, funding aimed at its expansion in Japan — building on a June 2026 partnership with JRE Ventures — and deeper investment in financial services capabilities. The company says it is trusted by global enterprises including MS&AD, Toshiba, PwC, KPMG, Nippon Paint and Keppel, and has processed over 1 billion files across finance, insurance, supply chain, healthcare and core business operations.

What the announcement does not say

As with most strategic collaboration agreements, the substance to watch is in what is missing. The announcement discloses no deal value, names no launch customers for the new agent suite, and sets out no timetable beyond bringing the agents to customers across Asia and the US faster through AWS and AWS Marketplace. The test of the deal is therefore concrete and observable: whether Marketplace distribution turns fileAI's existing enterprise footprint — 1 billion files processed — into production agentic deployments at banks and insurers, rather than more pilots. If the agents complete workflows with the audit trail the announcement promises, the “outcomes, not activity” slogan will have earned its keep; if not, it will read as exactly the kind of activity metric it criticises.

The original announcement is fileAI's blog post, with the PR Newswire release carried by TNGlobal.