Product Weekend Feb 2026 - Echo Series

The Feb Product Weekend is a little bit different from my usual ones.

Normally I spend the weekend building my own ideas. But this round I decided to try something else — pick up a few ideas from my clients and build quick prototypes for them.

Honestly, sometimes building based on real client needs is actually more interesting than just chasing your own ideas. You get clearer use cases, and you might learn something new along the way.

So that’s how the Echo Series came out.

Echo Desk

The first idea came from a client who wanted AI phone support.

That’s basically the core concept of Echo Desk — an AI that can answer phone calls and respond with voice using a provided knowledge base.

Most of the heavy lifting is actually handled by ElevenLabs. Their platform already supports voice agents, telephony integrations like Twilio, and even built-in RAG with uploaded knowledge.

So building the prototype was surprisingly straightforward.

One feature I experimented with is capturing unanswered questions. If the AI doesn’t know the answer, the system should log it so you can later update the knowledge base. Over time the agent should get smarter and more complete.

For testing, I used the built-in agent widget to simulate voice conversations.
Turns out the performance is actually pretty good.

Echo Map

The second idea is something quite different — a clickable map with AI voice responses.

The idea is simple:

You have a static map with different zones.
When a user clicks on a zone, the AI explains that location using voice and the supplied knowledge.

It’s almost like an interactive AI tour guide.

Among the three prototypes, this one is the most incomplete.

I managed to make it work for the first zone, but I got stuck when trying to pass the correct context to switch between different locations. The idea works, but the implementation still needs some work.

Still, it’s an interesting direction.

Echo Loop

The last one is actually my favorite.

Originally I thought Echo Loop wouldn’t involve voice at all.

The first prototype is basically an AI-assisted HR screening tool. You upload a bunch of resumes, define the job requirements, and the system rates each resume with a score. It can also generate follow-up interview questions.

That part was quite straightforward and the prototype came together pretty quickly.

Then my client suggested something interesting —
what if the system could automatically conduct a voice interview?

The idea is that after scoring candidates, the system schedules a voice call with the interviewee and asks the follow-up questions automatically, collecting their responses.

I haven’t built this part yet, but honestly it sounds both challenging and fun.

Wrap up

So yeah, that’s the Echo Series I built during the Feb Product Weekend.

Three quick prototypes:

• Echo Desk — AI voice phone support
• Echo Map — interactive map with AI voice guide
• Echo Loop — AI resume screening with voice interview idea

The initial feedback from the client is quite positive, and they’re already asking for enhanced versions.

I’m still figuring out whether any of these should become a real product… or just remain client prototypes.

But either way, it was a fun Product Weekend.