Problem

Accounting firms face a persistent capacity problem because skilled tax, audit, and accounting staff still spend substantial time on repetitive administrative work: gathering documents, searching client history, answering routine queries, preparing first drafts, and checking junior work. The pain is especially acute during tax season, when a year’s workload is compressed into roughly 12 weeks, causing burnout, client delays, and lost business when firms lack capacity. The clearest killer use case is tax-season execution: turning client correspondence and source documents into draft tax work and review-ready outputs, rather than having professionals perform every preparatory step manually.

The economics are primarily labor leverage and throughput. Combinely’s public customer example at Burgess Hodgson reports more than 2,600 tasks handled during December and January and a 75% reduction in the time required to create income-and-expenditure work. That points to a product aimed at freeing expensive professional capacity, allowing firms to handle more client work and reserve human attention for judgment, review, relationships, and advice.

Product / Service

Combinely is vertical, domain-specific AI for accounting services, positioned as an AI coworker or junior accountant that operates across tax, audit, and accounting workflows. Rather than replacing a firm’s existing software, its agents work across browser-based systems such as Xero, Outlook, and OneDrive, using firm knowledge and client context to perform repeatable tasks. The system is model-agnostic, routing work among models such as Claude, Gemini, and open-source models, and can be configured for tasks including tax research, client-response drafting, document extraction, reconciliations, payroll preparation, VAT checks, and first-pass review.

The delivery model is a secure, workflow-oriented platform accessed through the web app, email, or Microsoft Teams, with Combinely’s team helping create customized agents for specific firm processes. Firms can require proposed actions to be previewed and approved by a human, while agent activity is auditable; the company also reports SOC 2 Type II certification and says customer data is not used to train AI models. Pricing was reported as per seat, with the company using a demo-led sales process rather than publishing self-serve pricing. The intended benefit is to fit into existing ways of working while producing a reliable first version of the work for professionals to review and finalize.

Market

Combinely competes in vertical AI and enterprise workflow automation for accounting firms, spanning tax, audit, accounting production, and client-service operations. Its competitive set includes general-purpose assistants such as Microsoft Copilot and OpenClaw, traditional point solutions, and emerging accounting-agent companies such as Artifact AI, Archie, and Basis. Adjacent accounting automation products and platforms, including Karbon AI, Dext, AutoEntry, Hubdoc, Docyt, Vic.ai, and related tax and audit software, compete for portions of the same budget, although Combinely differentiates itself by automating cross-system, end-to-end firm workflows rather than a single data-extraction or accounting function.

The evidence indicates early commercial traction rather than a pre-revenue company. Burgess Hodgson, a practice with more than 150 staff, used Combinely for tax-season correspondence, draft accounts, and quality checks, while Blu Sky reportedly used it for more than 2,000 tasks in one month across VAT reviews, reconciliations, and Excel modelling. Combinely completed Y Combinator’s Spring 2025 program; PitchBook reports $500,000 of funding and deal records classified the company as generating revenue, while a third-party estimate put 2025 revenue at $440,000. Those figures should be treated as indicative rather than audited current financials, but the customer deployments and reported task volumes show that the product is being used in live accounting operations.