Avatar

Case – AI Assistant

2024–2025
The project has been running for 1 year and is not part of the NDA

One of my most rewarding projects at Bitsgap was the development of an AI-powered trading assistant. This project not only pushed the boundaries of what AI could do for cryptocurrency trading, but it also had a significant impact on the platform’s user experience and overall business outcomes.

AI Trading View

How We Built an AI Assistant for Traders at Bitsgap

The Problem

Traders feel overwhelmed. Despite using automated tools, many still rely on manual research — reading news, monitoring markets, and following influencers — just to feel in control of their trades. We believed AI could help.

Screenshot 1
Screenshot 2
Screenshot 3

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.

Step 1: User Research

We launched a survey to validate our assumptions.

1
Respondents
1,000+ users contacted, 258 responses received
2
What do they use?
18% said they read news daily, even while using automation
3
How many agreed?
40 users followed popular traders for insights
4
Conclusions
8% tracked Twitter, news channels, and forums — but only traded on spot, prioritizing safety
Survey Chart

We followed up with
10 in-depth interviews to understand

5
How they discover opportunities
6
What triggers them to execute a trade
7
Where they feel uncertain
Automation Insights

Step 2: Hypothesis Testing

To validate demand, we:

1
Testing the hypothesis
Created a landing page with a pre-order waitlist and real paywall (card pre-auth)
2
Result
Ran a targeted campaign → 58 paid subscribers ready to wait for the product
Hypothesis Test UI

This was a green light. But reality kicked in.

Step 3: Feasibility Check

Sentiment analysis = Expensive.
APIs (Reddit, Twitter/X, news feeds) cost a lot at scale.

After technical review, we pivoted

1
Used our own 5 years of trading data
2
Integrated backtesting engine
3
Focused on internal AI,
not external NLP APIs
Feasibility Summary

We updated our early subscribers and offered free early access to a new direction

Try our AI Trading Assistant — built with our proven data instead of news scraping.

80% of users opted in again and shared suggestions.

Step 4: Design & Prototyping

Our north star: zero-friction setup.

We designed:

1
One-click onboarding
2
3 risk levels: Conservative / Balanced / Aggressive
3
Pair recommendations and projected returns in USDT
4
Clear UI with probability ranges, not just “signals”
Step Flow
UI Screens

Step 5: Widget Design & System Integration

We built multiple widgets:

1
Strategy selector
2
Profit simulation
3
AI suggestions panel
3
Smart alerts & notifications

Internal QA and team testing showed promising results.

Step 6: Launch

1
Updated marketing site
2
Created onboarding email sequence
3
Added announcement banner on web app
Marketing Site Screenshot
Banner UI Screenshot

Step 7: The Outcome

The launch exceeded expectations:

1
80% early users joined the test
2
Increased subscription revenue
3
Growth in MAU and CRR
4
Positive user feedback + low churn
“It gave me confidence to trade without stress”

The AI Assistant didn’t just validate our hypothesis — it solved a real problem for cautious traders, without overcomplicating the experience.

Result

We found the sweet spot between:

1
User needs: clarity, safety, speed
2
Business goals: monetization, retention, differentiation

And we did it without burning money on expensive APIs.

MAU & CRR Chart
Subscriptions Chart