Overview
Bitsgap asked for an AI trading assistant that reduced research load without adding noise.
The work ran from hypothesis through survey, interviews, MVP, and launch — with one pivot when external data costs blocked the first direction.
Challenge
Traders already used automation but still spent hours on news, influencers, and manual research to feel in control. The product needed to replace overload — not add another feed.
Our Initial Hypothesis
What if we created an AI assistant that curates news and trading opportunities tailored to each user's "strategy" and "risk appetite"?
But First, We Needed Data
To validate our assumptions, we launched a comprehensive survey to understand how traders make decisions and how AI can help reduce "research overload".
Survey Results
We contacted 1,000+ users and received 258 responses. Here's what we discovered:
8% tracked Twitter, news channels, and forums - but only traded on spot, prioritizing safety.
Deep Dive: In-Depth Interviews
We followed up with 10 in-depth interviews to understand the deeper motivations and pain points of our users.
Where they feel uncertain and what causes the most anxiety in their trading.
From MVP to Real Users
With our research complete, we moved to the next phase: building and testing our AI Trading Assistant with real users.
Validating Our Hypothesis
To confirm our assumptions, we implemented a comprehensive testing strategy:
The Reality Check: "Sentiment Analysis" = Expensive
Our initial approach hit a major roadblock. APIs for Reddit, Twitter/X, and news feeds cost a lot at scale. After technical review, we had to pivot our strategy.
Instead of relying on expensive external APIs, we leveraged our internal resources and data:
Design and Prototyping
We started designing and prototyping with a clear north star: "zero-friction installation". We divided the process into 4 key steps:
Component Library: The Interface Behind the AI Assistant
We developed the design and integrated a comprehensive design system, creating several key widgets to enhance user experience.
We created multiple widgets to provide a comprehensive trading experience:
Product Launch
Yes, we successfully launched with internal quality control and team testing showing promising results.
The Launch Exceeded Expectations
Our AI Trading Assistant launch delivered impressive results across multiple key metrics:
Redesigned onboarding flow resulted in +11.3% user registrations.
AI trading assistant UX iterations achieved +16.8% task efficiency on key flows.
Deployed AI support chatbot reduced incoming support load by ~17.2%.
80% of early users joined the test, demonstrating strong product-market fit.
Key Learnings and Results
Our journey taught us valuable lessons about user needs and business goals: