The next phase of enterprise AI may depend less on better chatbots and more on what companies can do with the documents and unstructured data buried inside their existing systems. Singapore-based fileAI is targeting that problem with a new investment from SMBC Asia Rising Fund and Singtel Innov8, alongside the launch of its AI-native data-mapping product, fileScout.

fileAI Targets Enterprise Data Bottleneck With New SMBC and Singtel Investment

Enterprise AI adoption is entering a more difficult phase. Companies have moved from asking what generative AI can do to figuring out how to put it safely into finance, compliance, procurement, operations and customer workflows.

That shift is creating a market for technology that sits underneath the AI application itself: systems capable of turning contracts, statements, forms and other unstructured files into reliable data that enterprise software and AI agents can actually use.

Singapore-based fileAI, the company behind fileForge, is positioning itself in that layer. The company announced investment from SMBC Asia Rising Fund, the corporate venture capital arm of SMBC Group, and Singtel Innov8, Singtel Group’s corporate venture investment arm. The funding will support fileAI’s expansion in Japan and increase its investment in financial-services applications.

The financing is significant less for its disclosed amount—the companies have not publicly stated one—than for the strategic access it potentially gives fileAI to large, regulated enterprises.

SMBC brings relationships across banking and financial services, while Singtel provides another major corporate ecosystem in the region. For an enterprise AI company whose technology is designed for high-volume document and data workflows, those relationships could be as important as capital.

Turning unstructured information into operational data

fileAI’s core proposition is straightforward: enterprises have enormous quantities of information that conventional software cannot easily process because it is stored in documents rather than structured databases.

Its fileForge platform uses AI to capture, validate, match and reconcile information from complex files. The company says the resulting records can be delivered into enterprise systems as verified and audit-ready data.

The newly introduced fileScout focuses on the data-discovery and mapping problem. According to fileAI, it is designed to reduce the token consumption associated with processing unstructured enterprise data, before agent-based workflows perform tasks such as extraction, validation and reconciliation.

That approach places fileAI in a part of the enterprise AI stack that is becoming increasingly important.

The industry’s early focus was on foundation models from companies such as OpenAI, Google, Microsoft and Anthropic. Enterprises are now confronting a less glamorous constraint: those models are only as useful as the data and workflows surrounding them.

McKinsey’s June 2026 research on AI data readiness makes a similar point, arguing that companies scaling AI increasingly need governed foundations that connect structured and unstructured data.

For fileAI, that creates an opportunity to position document processing not as an isolated automation task, but as infrastructure for AI-powered enterprise workflows.

Japan becomes a strategic test market

The investment also accelerates fileAI’s push into Japan.

The company established a foothold there through a June 2026 partnership with JRE Ventures, the corporate venture capital arm supporting JR East Group. That relationship was aimed at using governed AI agents to digitize and structure legacy contracts and operational documents.

The latest investment is expected to help fileAI build a local Japanese team spanning sales, engineering and customer success.

Japan presents an attractive but demanding market for enterprise AI vendors. Large organizations often operate complex legacy systems and highly regulated workflows, creating demand for AI that can integrate with existing processes rather than simply sit alongside them.

For financial institutions, the potential use cases are particularly concrete. FileAI points to statement and covenant extraction, KYC and onboarding checks, reconciliation and regulatory reporting across multi-entity organizations.

These are not necessarily flashy generative-AI applications. They are the kinds of repetitive, document-heavy processes where even modest improvements in accuracy and processing time can have operational consequences.

Enterprise AI’s data problem is becoming a market of its own

The timing is notable because enterprise AI adoption is rising while production deployment remains difficult.

McKinsey’s 2025 global AI survey found that 88% of respondents said their organizations regularly used AI in at least one business function, yet most organizations were still experimenting or piloting at the enterprise level. Only about one-third reported that their organizations had begun scaling AI programs.

Asia-Pacific is moving in the same direction. IDC described 2026 as an “AI Supercycle” for the region, forecasting $110 billion in AI infrastructure spending by the end of the year and reporting that 75% of APJ organizations had deployed agentic AI in at least some initiatives.

But scaling AI requires more than compute.

IDC’s 2026 AI maturity benchmark found that only 3.1% of organizations worldwide had reached its optimized AI maturity stage, while 61.3% remained in the two least mature categories. The research identifies data quality, integration and production-grade platforms among the barriers holding organizations back.

That is precisely where vendors such as fileAI are trying to compete.

The broader competitive field includes document-intelligence platforms, intelligent process automation vendors, data integration companies and AI-native workflow providers. Larger technology companies including Microsoft, Google, Amazon and Salesforce can offer increasingly broad AI and data platforms, while specialist vendors can compete by going deeper into particular enterprise workflows.

fileAI’s differentiation is its focus on unstructured files, agent-based processing and the financial-services environments where accuracy, traceability and governance matter.

What enterprise buyers should watch

The investment does not by itself establish fileAI as a major global enterprise AI competitor. The more important test will be whether the company can convert its technology and strategic relationships into repeatable production deployments.

For enterprise technology teams, three issues will matter.

First is accuracy. Extracting information from documents is relatively easy when formats are standardized; maintaining accuracy across inconsistent, multilingual and legacy documents is substantially harder.

Second is governance. Financial and regulated enterprises need provenance, validation and auditability, not simply plausible AI-generated outputs.

Third is integration. An AI system that identifies information but cannot reliably push verified records into ERP, CRM, banking or workflow systems creates another silo rather than eliminating one.

The same challenge faces much of enterprise AI. Gartner said in June that more than one in 10 enterprises will be AI-first by 2030, with AI agents, semantics and converged data-and-analytics platforms among the technologies driving that transition.

fileAI’s bet is that the companies reaching that AI-first state will need an infrastructure layer capable of making their existing information usable.

The investment from SMBC and Singtel gives that thesis a stronger regional foundation. The harder part now is proving that unstructured-data intelligence can move from promising demonstrations into dependable, high-volume production workflows across Asia’s banks and large enterprises.

Market Landscape

Enterprise AI is increasingly developing into a layered technology stack:

• Foundation models: OpenAI, Google, Anthropic and other model providers supply general-purpose intelligence.
• Cloud and infrastructure: Microsoft Azure, Google Cloud, AWS and NVIDIA provide the compute and deployment infrastructure.
• Enterprise applications: Salesforce, Microsoft and others are embedding AI agents into business software.
• Data and integration: Data platforms and integration vendors connect AI to governed enterprise information.
• Document intelligence: Specialists such as fileAI focus on extracting and validating information from unstructured files.
• Agentic workflows: The emerging layer turns processed data into actions across finance, compliance, procurement and operations.

The strategic shift is from AI experimentation to AI industrialization. IDC says Asia-Pacific enterprises are moving toward AI operating models built around scalable data foundations, repeatable delivery and governance rather than isolated pilots.

That makes the unstructured-data layer increasingly important. Enterprises cannot simply deploy an AI agent on top of poorly organized documents and expect reliable production outcomes.

Top Insights

• fileAI secured backing from SMBC and Singtel while launching fileScout, strengthening its position in enterprise AI data infrastructure across Asia.
• fileScout targets unstructured enterprise data by reducing processing costs and preparing information for AI-driven capture, validation, matching and reconciliation workflows.
• Japan becomes a priority expansion market, with fileAI planning local sales, engineering and customer-success capabilities supported by SMBC’s enterprise network.
• Financial services provide a major target market, including KYC, regulatory reporting, covenant extraction, reconciliation and document-heavy onboarding workflows.
• The investment reflects a broader enterprise AI shift toward governed data foundations, as companies move beyond pilots and begin integrating AI into production systems.

Get in touch with our fintech expert