Tech news: Makosi rolls out AI agent for EBP audits

Makosi, an accounting and auditing solutions provider, announced the launch of its patent-pending AI agent EBP Eddy, made specifically for employee benefit plan audits. Makosi trained the generative AI agent on thousands of hours of technical training, regulatory information, industry best practices, and Makosi's Center of Excellence methodology. This enables EBP Eddy to function as a coach, copilot and creator. The AI helps users onboard new team members, answers EBP-related questions, and supports ongoing development, which is meant to be especially helpful for those new to EBP auditing; works alongside Makosi's auditors in real-time, offering contextual support such as help with working paper preparation, highlighting exceptions, validating completeness, drafting follow-up emails and sampling assistance; and builds efficiencies by assisting with data standardization, performing routine testing, and completing first-pass reviews of workpapers. Eddy is trained using Makosi's proprietary audit methodology, operates securely inside the firm's systems, and communicates via Teams, email, and voice. Built on Microsoft Azure infrastructure, the AI is ISO27001 certified and GDPR compliant. The solution does not require implementation, nor does it require firms to change systems, retrain staff, or manage another platform. It is already part of Makosi's delivery model.

Corporate accounting solutions provider OneStream announced the release of SensibleAI Agents, SensibleAI Studio and SensibleAI Account Reconciliations and the most advanced SensibleAI Forecast. 

SensibleAI Agents are AI-powered chat assistants that retrieve data, do deep analysis, visualize data and execute tasks. Embedded directly into the platform, SensibleAI Agents can be operated within OneStream workflows, data models, and security frameworks. They include a Finance Analyst Agent, which allows users to query financial data using natural language and receive real-time analysis, dashboards, and visualizations based on OneStream's financially intelligent engine; an Operations Analyst Agent, which provides on-demand access to operational and transactional data, such as inventory, transactions, and sales orders; a Search Analyst Agent which searches across enterprise documents, platform guides, and any unstructured data to transform information accessibility with source transparency; and a Deep Analysis Agent, which processes large volumes of documents to answer high-stakes, high-complexity questions. 

SensibleAI Studio is a growing library of more than 30 plug-and-play, quantitative and generative AI routines that users can activate within the OneStream platform or directly through APIs. Users can explore and analyze data in plain language and streamline reporting, reconciliation, and decision-making. Additionally, SensibleAI Studio's Genesis Blocks make it easy to plug AI capabilities into new or existing workflows, all while maintaining full transparency and auditability. SensibleAI Account Reconciliations is powered by SensibleAI Studio. It allows users to leverage AI to detect anomalies and flag risks—like unusual adjustments or missing documentation—early in the close process.

Finally, SensibleAI Forecast is a no-code, AI-powered forecasting solution embedded in the OneStream platform that brings together the company's machine learning-based forecasting engine (SensibleML) with new generative capabilities that allow finance teams to quickly model different scenarios and constraints, pulling forecasts up or down. Teams can create narrative commentary that provides context behind the forecast, strengthening analysis and decision making in real time.

Client collaboration platform Suralink announced the release of its new product Workpaper Suite and a collection of new AI automation and collaboration capabilities called Assessment Hub.

The Workpaper Suite is connected with the company's Request List Management solution, and brings the company's cloud platform and collaboration capabilities directly into desktop Excel. The solution provides automated document syncing between requests and workpapers to ensure workpapers have the latest client-provided information; pinned comments that can be created directly within Excel-based workpapers; cloud-based document storage so that Excel files don't crash, aren't downloaded onto desktops, and don't need to be manually deleted within each workpaper at the end of an engagement; and AI-powered document analysis. 

Meanwhile, the new Assessment Hub sports new capabilities within its flagship Request List Management product, which include the ability to quickly preview and search files without downloading them so users can instantly filter for newly uploaded documents, search and highlight matched text, and navigate across requests, all in a single view; pinned comments for smarter collaboration that let users add comments and tag firm or client users directly on documents, whether they're PDFs, spreadsheets, or images; AI-powered "ask anything" search that lets users search documents using natural language queries to locate details like invoice numbers or contract terms; and an enhanced user interface. 

Workpaper Suite and Assessment Hub are currently in early access and will be available for firms to start using in the next few weeks.