DEV Community

Baint Computer
Baint Computer

Posted on

Week 5 Building BAINT AI: When Users Don’t Think Like You

This week changed how we think about building BAINT AI

Not because of a new feature
But because of users

The Reality Check
We tested the BAINT AI

Classroom Assistant with real students.

And the feedback was immediate:
One user said:

“I was confused using it.”

Another said:

“I don’t face challenges in my studies.”

Two completely different perspectives

Same product.

The Problem Isn’t Always Technical

At first, it’s easy to assume:
The AI needs improvement

The system needs more features
But that wasn’t the case.

The real issue was:
User experience and expectation mismatch

We built a structured flow:
Pick a subject
Choose a topic
Ask questions

But some users expected:
Just type and ask immediately (like ChatGPT UI flow)

Two Types of Users
This week made something clear.

There are at least two types of users:

  1. Structured learners
    Prefer guided flow
    Want step-by-step progression

  2. Instant users
    Want quick access
    Prefer free-form input
    Designing for both is not simple.

What We’re Changing
Instead of choosing one direction,

we’re exploring a hybrid approach:
“Ask anything” (simple entry point)
Structured learning (guided experience)

The goal:
Reduce friction at the start, increase depth over time

Small Feedback, Big Impact
We also tried something different—reaching users in unexpected places

We asked a simple question in a live chat:
“What subject is hardest for you?”

One reply:
“Writing.”

That single word is now insight.

What We Learned
Users don’t think like builders

Clarity matters more than features

One piece of feedback can shape direction
Distribution is harder than building

Still Early
We’re still in the demo phase.
Still testing,adjusting,learning from every interaction

Final Thought
Building an AI product is not just about intelligence
It’s about making that intelligence accessible.

Top comments (0)