By Theodore Gavriilidis, Senior AI Engineer · 5 min read
Anthropic recently published a report analysing how people are building and deploying autonomous agents in production. Software engineering accounts for most of the usage. But an underrated detail is that back-office automation is the second largest category at 9.1% of all tool calls and it’s growing fast.
This tracks with our experience from building back-office automations for the past 2 years. A huge amount of back-office work in credit and financial services comes down to routine operations over unstructured documents: loan agreements, financial statements, due diligence packs, regulatory filings. The documents are messy, the formats change constantly and the stakes are high. For years the answer was either manual review or traditional IDP pipelines that needed months of configuration and broke the moment something looked different.
That era is ending. Agentic workflows are replacing the landscape and the shift is happening faster than most people in credit markets realise.
Figure 1: Source Anthropic
Why this matters for credit professionals
The old approach to document automation was a black box. Feed documents in, get outputs out. When something changed the process would usually fail and you’d need to spend time debugging the automation.
The shift we’re seeing now is toward what we think of as a glass box. Every step is traceable. Every extraction is sourced back to the original document. Low-confidence results get flagged for human review. The system does the heavy lifting at scale, but your team stays in control where it counts.
This is the right architecture for financial services: semi-autonomous, low-risk and auditable. Agents that can process hundreds of files, combined with structured review workflows where analysts approve, correct or escalate with full context and traceability.
What we’re building at Accuria
Our core focus at Accuria is building the infrastructure for this shift, specifically for credit professionals managing portfolios, transactions and reporting across asset classes and geographies.
AI Document Analysis is our document processing engine for credit work. It automates extraction pipelines that identify and classify loan contracts and transaction documents , flag ineligible documents in loan sales, extract financial covenants from complex agreements and process financial statements at scale. Every result traces back to its source. Every extraction carries a citation and a confidence score. Everything flows into review workflows where your team can efficiently verify outputs. This powers due diligence automation, covenant monitoring, regulatory compliance and report drafting. The core back-office workflows in credit markets.
AI Data Agent takes this further. It lets users get actionable insights from their structured and unstructured data in real time, through agents with domain knowledge built in. Need to stratify a portfolio, surface early warning signals, build a valuation or a cash flow model or onboard complicated source data into a target schema? What used to take hours of analyst work becomes a structured conversation.
The opportunity
The convergence of intelligent document processing, traceable human-in-the-loop review and domain-aware data agents is changing what’s possible in back-office operations. The traditional IDP industries are becoming obsolete. The teams that build on this infrastructure early won’t just cut costs. They’ll move faster, make better decisions and scale without proportionally scaling headcount.
We’re building this at Accuria for credit professionals globally, currently supporting €110 B+ in managed assets, performing and non performing loans, across 28 countries. If this is the problem you’re trying to solve in your organisation, reach out at accuria.com or talk to us.
Download the article Here
This presentation and the information contained herein does not constitute or form part of any (i) offer or invitation or inducement to sell or issue, or any solicitation of any offer to purchase or subscribe for, any securities or (ii) invitation or inducement to engage in investment activity within the meaning of Section 21 of the United Kingdom Financial Services and Markets Act 2000, as amended, nor shall any part of this presentation nor the fact of its distribution form part of or be relied on in connection with any contract or investment decision relating thereto.
