Reusable AI

Reusable AI

How Intuit is using reusable capabilities to scale AI

Intuit is a?global financial technology platform?serving over 100 million customers worldwide. Its offerings—TurboTax, CreditKarma, QuickBooks, and Mailchimp—are among the most powerful and intuitive tools aiming to solve the most important financial challenges for consumers and small businesses.

All the work we do at Intuit starts with our mission of powering prosperity around the world. Across our platform, we put the power of technology and data on the side of our customers, delivering simple, delightful solutions across the full range of our customers’ financial lives.

Using AI at scale, we automate and personalize their experiences, helping them make more money with less work while giving them full confidence in their actions and decisions. In fact, today AI drives 730 million customer help interactions per year and 58 billion machine learning predictions per day across our portfolio of products.

To scale AI at Intuit, we’ve adopted a reusable mindset across teams and embarked on a development path with self-serve Reusable AI services (RAISE) and Reusable AI Native Experiences (RAIN) that are enabling product teams to deploy machine learning (ML) capabilities without requiring deep interaction with AI+Data.

For example, Financial Error Detection Service (FEDS) is a RAISE capability that allows product owners to build AI-driven error detection models automatically via configuration. This speeds up the integration of new product features to serve use cases at scale by reducing the time and effort required when compared to a bespoke model.

One such use case is an invoice anomaly detection model, which understands the typical values entered by the user, alerts them of potential errors, and enables them to fix the errors, either immediately or later.

Write4U is an example of a RAIN capability. It’s a plug-and-play capability for autocomplete, powered by AI. It helps our customers finish their invoice descriptions, estimates, and transaction categories to reduce repetitive tasks for customers in QuickBooks Online, QuickBooks Live and QuickBooks Online Accountant products, and more. For an Intuit developer, using Write4U speeds integration time from months to hours, putting autocomplete capabilities at their fingertips. Switching to a different autocomplete model in their product becomes as easy as a configuration.

Our commitment to investing in building robust and reusable capabilities makes it possible for our development teams to deliver smart product experiences with speed, at scale. With RAISE, AI-driven product features that used to take months can now be implemented in a matter of days or hours.

If this blog has sparked your interest in how we’re innovating with AI and data here at Intuit, we’re always looking for great talent to join us. To learn more visit our?career website?and join our talent community!

Author:

Neha Malik, Product Manager

A seasoned Product Manager with over 15 years of experience in the US banking and financial services space, the last 6 years driving digital product initiatives. Currently, she holds the position of Product Manager at Intuit, where she is focused on the QuickBooks Online Payroll product. She is passionate about creating user-centered products that meet the needs of customers and drive business growth.

Co-author:

Tushar Manu Verma, Product Manager

Strategic and customer-obsessed product leader with 13 years of experience driving growth, improving retention, optimizing cost, and expanding market share with innovative SaaS & Desktop products across the Enterprise, Mid Market & SMB space. Specialization in Product Management, Consulting & Business operations across the domain of Fintech, Logistics & Manufacturing.

Prashanth Seshadri

Director AI, Analytics & Data. Hiring Data Scientists, ML Engineers, Data Engineers & Technical Analysts

1 年

Beautiful article and great examples ????

Lokesh Tripathi

Product Design | Product Management

1 年

Great article Tushar and Neha. Very insightful.

要查看或添加评论,请登录

社区洞察

其他会员也浏览了