AI for Audit and Compliance Practices_

Audit and compliance teams use AI to automate the collection and organisation of audit evidence from client systems, generate workpapers from trial balance data and supporting documentation, perform variance analysis and analytical review procedures, and manage the audit workflow from planning through fieldwork to final sign-off. The audit process is structured, evidence-driven, and documentation-heavy, making it a strong fit for AI assistance at every stage.

Why audit is ripe for automation

Statutory audit follows a defined methodology. ISA standards (International Standards on Auditing) prescribe planning procedures, risk assessment, substantive testing, and reporting requirements. Each audit engagement follows the same structure: accept and plan, assess risk, design responses, execute procedures, evaluate evidence, and form an opinion. The intellectual work is in judgment and evaluation. The mechanical work, gathering evidence, populating workpapers, performing calculations, and documenting procedures, consumes the majority of team hours.

Evidence gathering is the largest time sink. For a standard SME audit, the team might request and organise 200-500 individual documents: bank statements, sales invoices, purchase invoices, contracts, payroll records, loan agreements, fixed asset registers, and board minutes. Each document must be obtained from the client, filed in the correct audit area, and linked to the relevant workpaper. This process typically takes 2-4 days per engagement.

Workpaper preparation is the next bottleneck. Lead schedules, substantive testing workpapers, analytical review procedures, and completion checklists all follow standard templates but require manual population from the trial balance and supporting evidence. A standard SME audit generates 50-100 workpapers. Populating these manually takes 3-5 days of team time.

Analytical review, while analytically important, involves straightforward calculations: year-on-year comparisons, ratio analysis, trend analysis, and variance investigation. These calculations are repeated across every audit area and every engagement.

Use cases we build

Automated evidence gathering

AI connects to client accounting systems (Xero, Sage, QuickBooks, IRIS) and extracts the data and documents needed for each audit area. Bank statements are downloaded and matched to the trial balance. Sales and purchase invoices are sampled based on the audit plan’s sampling methodology. Contracts, loan agreements, and other supporting documents are requested from clients through an automated portal.

Each document is classified by audit area (revenue, purchases, payroll, fixed assets, cash, debtors, creditors) and mapped to the relevant assertion (existence, completeness, accuracy, valuation, rights and obligations). The audit team receives organised evidence rather than an unsorted pile of documents.

Gathering time drops from 2-4 days to 4-8 hours. The team starts substantive work faster rather than spending the first week of fieldwork chasing documents.

Typical timeline: 6-8 weeks. Typical investment: £15-25k / $20-30k.

Workpaper generation

AI populates workpaper templates from the trial balance, prior year audit file, and current year evidence. Lead schedules show current and prior year balances with variance calculations. Substantive testing workpapers are pre-populated with sample selections and relevant evidence. Analytical review workpapers show year-on-year movements, ratio analysis, and flagged variances that exceed materiality or expectations.

The auditor reviews pre-populated workpapers, confirms or adjusts the AI’s preliminary analysis, and documents their conclusions. The mechanical work of populating templates and performing calculations is handled before the auditor touches the file.

For a standard SME audit, workpaper preparation time drops from 3-5 days to 1-2 days of review and completion.

Typical timeline: 8-10 weeks. Typical investment: £20-35k / $25-45k.

Variance analysis and analytical procedures

AI performs analytical procedures across every audit area: calculates year-on-year movements, computes financial ratios (gross profit margin, current ratio, debtor days, creditor days), identifies trends over multiple periods, and flags variances that exceed materiality thresholds or expected ranges.

For each flagged variance, the system suggests potential explanations based on other data in the file (new revenue streams, cost increases, one-off transactions) and identifies the evidence needed to corroborate or refute each explanation. The auditor evaluates the explanations rather than performing the calculations manually.

Typical timeline: 4-6 weeks. Typical investment: £10-18k / $13-23k.

Audit planning and risk assessment

AI generates the audit planning document from prior year files, current year management accounts, and industry data. It assesses inherent risk by audit area, identifies significant risks, and proposes an audit approach with appropriate responses. The engagement partner reviews and adjusts the plan rather than drafting it from prior year notes.

The system also generates the audit strategy, calculates materiality (overall, performance, and trivial), and produces the team briefing document.

Typical timeline: 5-7 weeks. Typical investment: £12-20k / $15-25k.

Completion and reporting

AI generates the completion checklist, summarises unadjusted differences, prepares the management letter with control observations, and drafts the audit report. Going concern assessments are supported with cash flow analysis and ratio trends. Subsequent events review is prompted by monitoring Companies House filings and news sources for the client and its industry.

Typical timeline: 4-6 weeks. Typical investment: £10-18k / $13-23k.

How Formulaic approaches audit

Audit AI demands rigorous documentation. ISA 230 requires the auditor to prepare documentation that enables an experienced auditor with no previous connection to the engagement to understand the work performed. Every AI-generated workpaper meets this standard by showing its data sources, calculation methodology, and the basis for any preliminary conclusions.

We integrate with your existing audit software. CaseWare, Inflo, and Pentana integrations handle workpaper filing and review workflows. Xero, Sage, QuickBooks, and IRIS connections handle evidence extraction. The AI sits between your audit methodology and your client’s records, filling the gap that currently consumes team hours.

Quality control workflows are built in. Partner and manager review points are maintained. The AI doesn’t replace the review process; it ensures that what reaches the reviewer is complete and well-documented.

Data security meets the requirements of your professional body (ICAEW, ACCA, ICAS) and your clients’ expectations. Client data stays on UK-hosted infrastructure with encryption, access controls, and retention policies.

We start with the audit: £3,500 / $4,500 over two weeks to assess your current audit methodology, software stack, and team workflow. The output is a build plan prioritised by time savings per engagement.