MY ROLE

I was responsible for the end-to-end design of the new Insights Assistant feature for the generative AI SaaS platform. I conducted user research to identify needs and pain points, developed concepts in collaboration with cross-functional teams and created wireframes and interactive prototypes. I led usability testing to refine the user experience and designed the interface to align with Kaitongo’s brand. Working closely with developers, I ensured seamless implementation and gathered post-launch feedback to guide future iterations.

OVERVIEW

The Problem

Kaitongo, initially a news platform with advanced search and tagging, needed to add value and drive revenue. Stakeholders wanted the platform to be more proactive and generate actionable insights. The challenge was to understand the users’ workflow and identify how Kaitongo could better serve them post-search.

The Outcome

I designed the Insights Assistant feature, which transformed Kaitongo into a proactive tool for generating actionable insights. By simulating junior analysts’ workflows with GPT agents, we created an intuitive, efficient experience. This feature led to three new client acquisitions and significant praise for its ability to streamline users’ research and decision-making processes.

THE CHALLENGE

Complexity

Designing the Insights Assistant required deeply understanding analysts’ complex workflows. I translated these processes into clear journeys to communicate effectively with the machine learning team. Ensuring the new experience felt natural and seamlessly integrated into users’ routines was crucial.

Too Many Stakeholders

I dealt with diverse stakeholders, including investors, founders, machine learning engineers, existing users, developers, and the marketing team. With multiple opinions but no research-backed decisions, balancing these perspectives was challenging. Ensuring existing users wouldn’t feel lost or overwhelmed was also essential.

Time + Budget

With a tight deadline of 6 weeks and a constrained budget, we had to allocate resources efficiently and prioritize tasks effectively. Rapid iteration and quick turnarounds were essential to meet our goals without compromising on the quality of the Insights Assistant feature.

Omni-channel Touchpoints

The workflow we aimed to mimic involved multiple omni-channel touchpoints within the firm. I had to study every interaction within a team to accurately replicate the workflow, ensuring a seamless and cohesive experience across all channels.

New To Market

Introducing a novel feature like the Insights Assistant meant navigating uncharted territory. We had to ensure the feature addressed real user needs, stood out in a competitive market, and demonstrated clear value, all while dealing with the uncertainties inherent in launching a new product.

Over-Support Issues

With extensive teams in sales and customer success, there was excessive hand-holding. Kaitongo suffered from over-support issues, which made the leadership blind to the core issues related to discoverability, usability, or significant pain points in the product.

THE APPROACH

Build it once, perfect it endlessly.

Key Features

  • Quick validation of design
  • Continuous user and stakeholder feedback
  • Ability to test against fixed outcomes
  • Faster ship times
  • Assessing customer satisfaction by rapid prototyping

Obstacles

  • Team alignment can be tricky
  • Absence of UX research before sprint 0
  • Too many responsibilities to navigate