What problem led you to start Eclatira, and when did you realize it could become a company?
Moad Rahali Semlali: Before Eclatira, we worked as a service provider and spoke with more than 100 business owners. Most of them asked us for the same thing: an AI voice agent that makes sure no call is ever missed and every client request is captured. None of them wanted to hear “I called you, but no one picked up” again.
We built those solutions on the AI voice platforms available at the time, and the results fell short. The owners told us plainly: “This is not the quality of AI I can put in front of my customers.” The problems were latency, hallucinations, strange behaviour, and inconsistency from one call to the next.
That was the turning point. In September 2025, we decided to re-engineer the technology ourselves instead of building on what the market offered. Most platforms chain speech-to-text, a language model, and text-to-speech together, and every handoff adds delay and room for error.
Eclatira runs on a speech-to-speech conversational engine and a real-time voice and video protocol we engineered ourselves. That architecture solved the issues that made AI voice agents unusable for those owners, and it’s what turned Eclatira from a project into a company.
Eclatira’s agents can respond to what a person says and shows them. What does that make possible that a voice-only agent cannot do?
Moad Rahali Semlali: It delivers the kind of support you would get from a human assistant who is actually looking at your problem. A voice-only agent has to guess from a description. An Eclatira agent can see.
- Guided screen support: A customer shares their screen, and the agent sees the error or the form they’re stuck on and walks them through it step by step.
- Product and item inspection: A customer shows a damaged product, a document, or a label on camera, and the agent understands it right away.
- Onboarding and sales demos: The agent guides a prospect through a product live and sees exactly where they hesitate.
- Healthcare and pharmacy intake: A patient holds up a medication box or an insurance card, and the agent captures the details accurately without spelling anything out over the phone.
This is the step from AI voice agent to real-time conversational video AI: an assistant that listens, sees, speaks, and takes action in your systems, all in the same conversation.
Who are Eclatira’s clients today? Can you share an example of how one uses the platform and the results they’ve seen?
Moad Rahali Semlali: Eclatira has paying customers in hospitality, pharmacy, legal services, and e-commerce. They use the platform as an AI receptionist, an AI answering service, an AI lead qualification agent, and an AI customer service agent.
One example is a hospitality booking platform that moved its AI agent to Eclatira from another voice AI provider. Its Eclatira agent handles reservation inquiries across channels, answers guest questions, and connects directly to its booking systems through automation, so every request is captured and acted on without staff getting involved.
The biggest change the team noticed was quality: guests now get natural, consistent conversations instead of the robotic delays that made the previous system hard to trust. The front desk no longer loses requests after hours or at peak times.
Which customer problem are you most focused on solving first, and what have you learned from people using the platform?
Moad Rahali Semlali: There are two, in this order. First: no missed call, no lost request, and no unanswered customer, 24/7. Second: doing it at a quality level business owners are proud to put in front of their customers.
Our biggest lesson is that business owners don’t judge AI on features; they judge it on trust. One awkward pause, one invented answer, or one inconsistent response is enough for them to turn it off.
Latency and reliability aren’t technical details to them; they decide whether the business adopts AI at all. That’s why we built our own real-time engine instead of stitching third-party tools together.
When an AI agent can see a camera feed or shared screen, how do you address privacy, accuracy, and the need to hand a conversation to a person?
- Privacy: Camera frames are never stored and never used for training. They are streamed live to the AI model for that conversation only. Customers turn on video within the same session, directly in the browser, with nothing to install.
- Accuracy: Eclatira agents answer from the business’s own knowledge base and connected systems, not from general guesses. Each agent is scoped to its specific use case, which keeps it focused and consistent.
- Human handoff: When a conversation needs a person, the agent transfers it live to a staff member. The AI handles the volume, and people handle the moments that need them.
What has been the hardest part of building Eclatira, and what milestone are you working toward next?
Moad Rahali Semlali: The hardest part was deciding to rebuild the stack instead of assembling existing tools, and then making that engine production-grade.
The second hardest was real-time synchronization. Early on, if a user held something up to the camera, the agent registered it a few seconds late. In a live video conversation, that breaks the experience, so we re-architected how visual context reaches the agent. Now it reacts to what it sees as it happens.
As for what’s next, we have already opened Eclatira to developers. Teams can plug real-time conversational voice and video AI into their own apps through our API, SDK, and MCP support, with access to more than 3,000 app integrations through automation platforms such as Make.com and n8n.
At our Product Hunt launch in September 2026, Eclatira ranked #2 among AI agents and assistants. Our next milestone is to make Eclatira an essential tool for every developer building AI copilots, video-first AI products, and voice agents.
About Eclatira
Eclatira is a Quebec-based Canadian AI company co-founded by Moad Rahali Semlali, CEO, and Maryam Gilsenan, CTO. It builds a real-time conversational voice and video AI platform for business customer service and developer applications.
To explore the platform, visit Eclatira, learn about its AI receptionist and live video AI, or browse its developer documentation.
