Indigo AI Smart Assistant
Finance doesn't have to be complex.
True Potential LLP
Finance, Technology, AI
Lead Product Designer
3 months
Setting the scene
Financial advisers typically spend a lot of time on admin. Writing emails, pulling together fact finds, answering repeat questions, documenting calls… the list goes on. It eats up hours that could be spent doing what advisers want to do - spending time in front of their clients.
That’s where Indigo AI came in. Our idea: build a smart assistant that lives inside the adviser platform, automating the boring stuff, speeding up the day-to-day, and giving advisers back their time.
The goals
Save advisers time by automating routine tasks.
Boost confidence by giving instant, accurate answers to queries.
Unlock more revenue - less admin = more face time with clients.
Lower business costs by reducing demand on support staff.
Be versatile enough for staff across the company, not just advisers.
This tool could reach 20% of all UK financial advisers.
The challenges
Trust: Advisers worried - is the data secure? Is it accurate? Could mistakes cost clients financial harm?
Infrastructure: AI queries big data. Bad infrastructure = long waits = frustrated users.
Adoption: Average adviser age is 56. AI might sound “too techy” unless the UI feels approachable.
We couldn’t just build something powerful. It had to feel safe, simple, and smart.
Research & insights
User interviews
I teamed up with a colleague to interview a wide range of people: advisers, case checkers, account managers, marketing staff, and investment managers.
What we found:
Advisers wanted help drafting emails and creating client-ready documents.
Account managers wanted to know where advisers were in the pipeline.
Case checkers cared about accuracy and speed in reviewing.
Marketing wanted summaries and content shortcuts.
Everyone wanted something that felt personalised to their workflow.
“ Many advisers regularly complete straightforward reports, such as Withdrawal Letters, GIA to ISA letters, and Fund Switches letters, by manually inputting data from their system into Word templates. An AI assistant could streamline this process by generating an initial draft report populated with client-specific information. ”
Tom R
True Potential Financial Adviser
Competitive benchmarking
AI products are everywhere - but most are either too generic (like ChatGPT) or too niche. No one was offering an integrated, fully locked-down AI assistant tailored to financial services.
I benchmarked popular conversational AI tools, made mood boards, and picked apart what worked and what didn’t. The insight? If we could marry the power of AI with a finance-specific workflow, we’d have something totally unique in the industry.
Mapping it out
Using Miro, I created user flows for each department using the product. That way, we weren’t just designing one generic assistant, we were designing personalised experiences that adapted to the user’s role.
For example:
Financial Advisers could generate a fact find or client summary.
Account Managers could get updates on submissions.
Marketing could create tailored comms.
Staff could generate internal documents quickly.
In the below example, after meeting with Account Managers who work with Advisers within the business, they highlighted the most common duties of their role:
Summarising an adviser
Checking for an update on a submission case
What stage an adviser is up to on their online courses / assessments
Viewing policy messages.
The aim of this would allow Account managers to manage their role more efficiently, this would not only save time for themselves, but have a knock on effect with the advisers they work with. Allowing them to achieve the main objective - spending more time in front of their clients.
From Idea to Interface
Sketching: Quick sketches to explore layouts before sinking time into visuals.
Wireframes: Low/medium fidelity mockups in Adobe XD, keeping things minimal and branded.
Interactive Prototypes: Clickable versions advisers and stakeholders could play with - perfect for testing flows early.
Feedback Loops: Internal testing with advisers and staff to refine.
High Fidelity Designs
Once flows and layouts were validated, I moved to polished UI. I combined our existing adviser design system with a new Indigo AI design system I built just for this project.
The look:
Clean, professional, but approachable.
On brand whilst exploring Indigo AI's own identity
Clear prompts and outputs.
Scalable components to make future features easier to add.
Testing & Feedback
We tested internally with advisers and staff. Key takeaways:
The assistant felt intuitive and easy to use.
Users liked how the AI could be tailored to their department.
Stakeholders saw clear time-saving benefits and supported rollout.
Wrapping Up
The product was launched in March 2026 and early feedback from the advisers was very positive.
Advisers used on launch day:
182
Conversations on this period:
378
Total prompts on this period:
832
“ I am absolutely buzzing - this is going to save so much time. Pass my thanks onto the guys for all this - its amazing.”
Michael M
True Potential Financial Adviser