Intuit Assist, AI help for tax experts on live calls
Intuit’s live experts, more than 15,000 across TurboTax, QuickBooks and Credit Karma, spend their day with customers who have complicated questions and real worry about money. We built Intuit Assist into the tools experts already use, to take on the note-taking, searching and follow-up that fill a shift. Every suggestion stays optional.
- +12
- Point lift in NPS
- 15 min
- Less handling time per call
- 88%
- Experts using it daily
- Role
- Sole designer, mentoring a junior designer
- Worked with
- Engineering, content, brand, design systems, product leadership and Intuit’s AI organization
- Timeline
- 9 months
- I owned
- Strategy, research, prototyping, visual design, the design system, executive reviews and final QA
Decision 01
Design for the pain points across the whole journey
- Research
- Moderated sessions watching experts and customers in real time, plus interviews about daily frustrations. Privacy rules were strict, so we worked from anonymized journeys and aggregated insights.
- Situation
- Research turned up pain points across the whole customer and expert journey: calls getting longer as issues grew more complex, notes and follow-up emails eating into each shift, and experts who knew the tax language but wanted help writing with clarity and empathy.
- Choice
- Instead of one assistant feature, a set of features that take over or ease those pain points before, during and after the call, starting with support calls.
- Why
- Each pain point pulled the expert’s attention away from the customer, so each one got its own fix.
Decision 02
Keep customer data on screen
- Situation
- Experts needed Intuit Assist within reach, but the screen was already full of customer information they relied on during a call.
- Choice
- A small entry button in the global header, chosen after several rounds of testing, and plain-language answers right in the panel, so experts never leave the customer’s screen to look something up.
- Why
- It keeps high-value information visible and the assistant one click away at any point in the call.
- Result
- In early testing, answering questions in place saved an average of twelve minutes per call.
Decision 03
Do the prep and the note-taking
- Situation
- Before a call, experts searched records for context. During it, they took notes by hand while trying to listen.
- Choice
- Intuit Assist summarizes the customer before the expert joins, then follows the conversation and turns it into structured notes.
- Why
- Experts could stay with the customer instead of the keyboard, and the notes carried over to the next interaction.
Decision 04
Make every suggestion optional and editable
- Situation
- In testing, experts wanted to review and edit generated content before it reached a customer.
- Choice
- Recommended replies that experts can edit to match their own tone, and follow-up emails drafted from the call with adjustable tone and length. Suggestions are always optional.
- Why
- Editing and approval controls increased trust in the output. After follow-up testing we added more editing controls, plus quick-select options for faster adjustments.
Decision 05
Let experts grade the output
- Situation
- The model needed a way to improve from everyday use.
- Choice
- Thumbs up and down on AI output, with a short list of reasons for a thumbs down: not factually correct, doesn’t make sense, generic, or too long.
- Why
- Expert feedback shapes future improvements to the model.
Decision 06
Earn trust before the first live call
- Situation
- Experts didn’t trust AI yet, and a new assistant in the middle of a live call with a worried customer is a lot to take in.
- Choice
- An animated onboarding that introduces what Intuit Assist can do, and a practice simulator where experts could try it on mock calls before using it with real customers.
- Why
- Experts learned the assistant at their own pace, with no customer waiting, so their first real call wasn’t their first try.
Decision 07
Build the missing components as a small design system
- Situation
- Intuit Assist brought patterns Intuit’s design system didn’t have yet. Intuit Assist had its own component library, but none of its components fit the Expert Portal, the dashboard I designed for Intuit’s experts. Because the portal isn’t customer-facing, the design system team didn’t support it.
- Choice
- I built the components myself: a small design system for the Expert Portal, based on Intuit Assist’s component library, with buttons, chips, cards, modals, avatars and AI panels.
- Why
- Every AI feature in the portal started from the same parts, so it looked and worked the same, with no design system team behind it.
Results
+12
point increase in NPS
15 min
less handling time per call, on average
88%
of experts used it daily after launch
28%
less silence and hold time per call
What I learned
- 1
Control drives adoption.
Early adopters liked the speed. Being able to edit and approve is what made them trust the output, so we added more of it after follow-up testing.
- 2
Writing help mattered as much as answers.
Experts already knew the tax language. Generated summaries helped them write clearer responses to customers.
- 3
Status matters after the call.
Experts asked for clearer signals when a summary was ready.
What experts wanted was a faster way to act on their own judgment. Intuit Assist gave them time back and let them sound more like themselves on calls.
