Intuit advances GenOS to accelerate agentic artificial intelligence development

Intuit expanded its GenOS platform with custom financial large language models, expert-in-the-loop tooling, and agent evaluation frameworks to speed agentic artificial intelligence across its products. Early results show improved accuracy and significantly lower latency in accounting workflows, with more agents rolling out soon.

Intuit announced a major update to its proprietary Generative Artificial Intelligence Operating System, GenOS, describing rapid progress over the past 90 days to scale done-for-you agentic artificial intelligence experiences across its financial technology platform. GenOS already underpins offerings such as the QuickBooks Online Virtual Team of artificial intelligence agents and new agentic capabilities in the Intuit Enterprise Suite. The company positions these systems to automate everyday tasks, manage complex workflows, and connect customers to both artificial intelligence agents and artificial intelligence enabled human experts, with the goal of driving better financial outcomes for consumers, small businesses, mid-market firms, and accountants.

The release adds three core advancements. First, custom-trained Financial Intuit large language models are fine-tuned on financial data and are already powering agentic features in QuickBooks Online and the Enterprise Suite. Intuit reports early gains of 5 percent higher accuracy and 50 percent lower latency for some accounting workflows versus certain general-purpose off-the-shelf models, with indications of cost efficiencies. Second, new expert-in-the-loop capabilities in GenUX let developers embed human expertise directly into agent workflows, with tools for intelligent routing, orchestration, and seamless handoffs to Intuit’s network of tax and bookkeeping experts. Third, enhancements to the Agent Starter Kit with a GenOS Evaluation Service provide methodologies, dashboards, and a toolchain for monitoring, iterating, and assuring agent performance at scale. Thousands of Intuit developers are using these ready-made components to build, test, and optimize artificial intelligence agent experiences.

Intuit’s chief technology officer, Alex Balazs, said the updated GenOS accelerates the company’s pace of innovation and advances its ambition to serve as a system of intelligence for customers. He highlighted early production results from the proprietary financial models in accounting workflows, describing the release as a step-change in delivering robust generative and agentic artificial intelligence experiences.

Looking ahead, Intuit plans additional artificial intelligence agents for consumers, businesses, and accountants. Upcoming capabilities include agents in QuickBooks Online tailored for larger and growing businesses, with the Payroll Agent and Project Management Agent already in beta. The company notes that three years of investment in GenOS have enabled rapid innovation for approximately 100 million customers and empowered thousands of internal technologists to design and deploy experiences responsibly. Intuit cites a large data foundation, including 625,000 customer and financial attributes per small business, 70,000 tax and financial attributes per consumer, and 60 billion machine learning predictions daily. The company underscores commitments to data privacy, security, and responsible artificial intelligence governance and is a member of the Center for Artificial Intelligence Standards and Innovation established by NIST.

70

Impact Score

Will agentic Artificial Intelligence disrupt SaaS?

Bain argues that agentic Artificial Intelligence is set to reshape software as a service by automating tasks and rebundling control across a new stack. The firm outlines four disruption scenarios and a playbook for incumbents on data, standards, pricing, and talent.

Advancing cutting edge artificial intelligence in health

Google details how breakthroughs like Gemini, Gemma variants, AlphaFold, and enterprise tools are being applied across healthcare. The company highlights research, open models, and clinical search products that aim to make care more personalized, accessible, and effective with Artificial Intelligence.

Open artificial intelligence models shown secure for enterprise adoption

A LatticeFlow evaluation with SambaNova found that targeted guardrails can lift security scores of leading open-source generative models from as low as 1.8 percent to 99.6 percent while preserving service quality. The findings position open models as viable for regulated sectors such as financial services.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.