Digits announces Autonomous General Ledger

Accounting automation solutions provider Digits announced the release of its new Autonomous Digital Ledger solution, directly challenging competitors such as QuickBooks.

The AGL acts as a proactive bookkeeper, offering real-time reconciliation, categorization and classification of inbound transactions and financial data. Deep learning enables it to detect recurring transactions, even upon their first arrival. Custom large language models parse through transactions, bank statements, contracts and invoices to give better context for classifications. The AI does constant variance analysis, which tracks momentum and magnitude of charges in order to promptly flag what matters. The product also features a large vendor data set that can be used to standardize client ledgers for year-end tax preparations.

It also continually learns, as Digits models user data in a vector database, which allows it to continuously index and search a user's entire financial history in real time and deliver insights. Meanwhile, AI agent computer programs that perceive and interpret their environment can autonomously perform actions and make decisions like identifying transactions for accruals and depreciation schedules and automatically creating entries and supporting documentation.

While AI-driven accounting isn't necessarily new, Digits noted that many of the programs rely on large language models like ChatGPT, which aren't specialized for precise mathematical work and represent possible data security concerns. Instead, the solution is built on dozens of specialized AI models and agents that were produced and trained in-house on a proprietary dataset of over $825 billion in small-business transactions. These models, among other things, are trained to find similar transactions or vendors to avoid the need to manually categorize every expense. This allows for pattern recognition and the ability to mimic the unique behaviors of each individual Digits user. Digits claims its proprietary models outperform GPT-4o by 54%.

"LLMs are an amazing technology, but they're also famously prone to hallucination," said Jeff Seibert, CEO and co-founder of Digits. "In accounting, that's unacceptable. At Digits, we've pioneered application-specific AI Accounting models to automate over 90% of SMB bookkeeping workflows, saving business owners and accountants countless hours every month."

Digits AGL is now available as self-serve software, with pricing starting at $100 per month for startups and small businesses. Expanded features for accounting firms will roll out later this year, with testing already underway with firms such as Armanino and Hiline.

With this launch, Digits also announced the appointment of industry veteran Craig Walker, co-founder and former chief technology officer of small business platform Xero, to its leadership team as product advisor. 

"Digits has fully harnessed the power of AI," he said. "The categorization is magical. What took me a week to set up in Xero or Quickbooks, took me just minutes in Digits—you truly feel the AI at work. Intelligent accounting software is now possible, and Digits is the first to get it right. If Xero was the pioneer of cloud accounting, Digits is the pioneer of AI accounting." 

The announcement comes just one month after Digits announced  the release of an AI-guided invoice automation solution. The system uses AI to generate invoices, pulling in customer details, pricing and payment terms with minimal user input. It keeps tabs on incoming payments and knows when invoices are viewed, paid or overdue with tracking and automated follow-ups (the AI automatically sends gentle reminders and escalates follow-ups when needed.) AI Invoicing integrates directly with the user's Digits accounting system. Users can personalize their invoices with their own logo, branding and custom fields. Digits also claimed "bank-grade security, including SOC 2 Type II compliance" to protect sensitive financial information within the Digits encrypted infrastructure.

For reprint and licensing requests for this article, click here.
Technology Bookkeeping Artificial intelligence Machine learning
MORE FROM ACCOUNTING TODAY