Enterprise AI is entering a less glamorous but increasingly important phase: making sense of the documents, files and fragmented data that sit underneath everyday business processes. Singapore-based fileAI is targeting that layer with fresh backing from SMBC Asia Rising Fund and Singtel Innov8, alongside the launch of its fileScout unstructured-data mapping technology.
The investment announced Thursday gives fileAI new strategic backing as it expands in Japan and builds deeper capabilities for financial services, while positioning its technology around a problem that is becoming harder to ignore: getting enterprise data into a form AI systems can actually use.
Financial institutions, insurers and large corporations may have sophisticated cloud infrastructure and access to increasingly capable large language models (LLMs), but much of the information those systems need remains buried in contracts, statements, forms, invoices, correspondence and other unstructured files.
That makes data preparation more than a back-office concern. It is becoming part of the enterprise AI stack.
fileAI, the company behind the fileForge platform, says its technology is designed to capture, validate, match and reconcile information from complex files before publishing structured, verified records into downstream enterprise systems. Its newly introduced fileScout focuses specifically on mapping unstructured enterprise data and selecting relevant information before extraction, with the goal of reducing unnecessary token consumption.
The distinction matters. Sending every page of a document corpus to a large model can be expensive and can introduce irrelevant context. A system that first maps documents, identifies useful fields and then routes information through extraction and validation workflows takes a more infrastructure-oriented approach.
That places fileAI in an increasingly crowded enterprise AI market, but not necessarily in direct competition with the largest foundation-model providers.
Companies such as Microsoft, Google, Amazon, Salesforce and Adobe are embedding generative AI into enterprise software, productivity suites, customer platforms and cloud infrastructure. Their advantage is distribution and access to existing business systems. fileAI’s proposition is narrower: act as an intelligence and data-processing layer for messy enterprise information that must be converted into reliable operational data.
That is a significant distinction for regulated industries.
In banking, for example, AI may need to extract information from financial statements, validate KYC documentation, reconcile records across entities or identify terms in lending agreements. The output cannot simply be a plausible answer from an LLM. It needs traceability, validation and a route into the systems responsible for the next business action.
fileAI says its platform combines agentified data capture, validation, matching and reconciliation with governed workflows. The company is also expanding financial-services capabilities through the new investment.
The strategic involvement of SMBC is particularly relevant here. SMBC Asia Rising Fund is the corporate venture capital fund of SMBC Group, established to invest in high-potential Asian startups and accelerate business development and partnerships. Its portfolio already includes enterprise AI and fintech companies.
Singtel Innov8 brings a different strategic connection. The corporate venture arm of Singtel Group invests in technologies spanning enterprise AI, cybersecurity, automation and digital infrastructure, while using Singtel’s regional footprint to connect startups with potential customers and partners.
For fileAI, that network may prove as important as the capital itself.
The company says the new funding will support a local Japan team spanning sales, engineering and customer success. The move follows a June 2026 partnership with JRE Ventures, the corporate venture capital arm supporting the JR East Group, which established an initial foothold for fileAI in Japan.
Japan is an important test market for enterprise AI vendors because large organizations often operate across legacy systems, extensive documentation and highly structured compliance requirements. Winning deployments there can require more than model accuracy; vendors need integration, governance and the ability to handle operational edge cases.
That challenge mirrors a broader shift in enterprise AI.
McKinsey’s 2025 global AI survey found that 88% of respondents said their organizations regularly used AI in at least one business function. Yet nearly two-thirds said their organizations had not begun scaling AI across the enterprise. Only 39% reported an enterprise-level EBIT impact.
The implication is straightforward: AI adoption is no longer primarily about convincing companies to experiment. The harder problem is building the infrastructure required to move successful experiments into production.
Unstructured data is a major part of that infrastructure problem. McKinsey has described unstructured information such as emails, contracts, conversations and documents as a growing obstacle to AI scaling because data can lose context as it moves through extraction, chunking and embedding processes.
This is where fileScout and fileForge are attempting to establish a foothold.
The company’s technology does not eliminate the need for foundation models. Instead, it sits before and around them, preparing information, applying schemas and validations, and turning model-assisted extraction into structured business records.
For enterprise buyers, that architecture could be more consequential than another incremental improvement in an LLM benchmark. The value proposition is ultimately about whether AI can operate reliably inside finance, procurement, compliance and customer operations without creating a new layer of manual verification.
There are still questions fileAI will need to answer as it scales. Enterprise customers will scrutinize model accuracy, integration depth, security, auditability, deployment economics and how the platform performs against established document-processing, intelligent automation and cloud AI offerings.
The competitive field also continues to evolve. Gartner has warned that enterprises face governance and security challenges as AI agents proliferate, with only 13% of surveyed organizations saying they had the right governance structures in place for agentic AI in 2025.
That makes the idea of a governed data layer increasingly relevant.
fileAI is betting that enterprises will need specialized infrastructure between their raw information and increasingly autonomous AI systems. The SMBC and Singtel investments give the company additional regional relationships from which to test that thesis.
The bigger story is not simply another enterprise AI funding announcement. It is the emergence of a new battleground in AI infrastructure: the systems that determine whether corporate data is clean, contextualized and trustworthy enough for machines to act on.
Market Landscape
Enterprise AI is shifting from chatbot experimentation toward workflow automation, data orchestration and agentic systems. McKinsey found that 88% of surveyed organizations were regularly using AI in at least one business function in 2025, but only a minority had begun scaling it broadly.
That gap creates opportunities for infrastructure vendors focused on data readiness, governance and workflow integration.
The competitive landscape spans several layers:
- Cloud AI: Microsoft Azure, Google Cloud and Amazon Web Services provide foundation-model access, data infrastructure and enterprise AI tooling.
- Enterprise software: Salesforce, Adobe and Microsoft are embedding AI directly into CRM, productivity, marketing and workflow applications.
- Document intelligence: Specialist vendors focus on extracting information from invoices, contracts, forms and other business documents.
- AI infrastructure: Newer platforms increasingly address retrieval, orchestration, governance, agent execution and unstructured-data preparation.
fileAI is positioning fileForge toward the intersection of these categories, with fileScout focused on mapping unstructured data before it enters downstream AI workflows.
For enterprise technology teams, the practical question is therefore less whether AI can read a document and more how reliably the resulting data can drive a business process. That shifts procurement priorities toward validation, auditability, integrations, cost control and governance.
Top Insights
- fileAI has secured strategic investment from SMBC Asia Rising Fund and Singtel Innov8, supporting Japan expansion and deeper financial-services AI capabilities.
- Its new fileScout technology maps unstructured enterprise data before extraction, potentially reducing token costs while improving downstream AI workflow efficiency.
- fileForge targets document-heavy enterprise processes including finance, compliance, procurement and onboarding, where accuracy and auditability matter as much as model capability.
- The announcement reflects a wider enterprise AI shift from pilots toward production workflows, where data readiness and governance increasingly determine business value.
- SMBC and Singtel provide fileAI with strategic enterprise ecosystems across banking, telecommunications and Asia-Pacific markets, potentially accelerating customer adoption beyond funding alone.
