Lautaro Nievas

When users share key info (like their name, contact reason, or preferred time), the assistant keeps that in context. That way, it won’t ask again five minutes later. This creates more natural conversations, reduces steps, and makes the assistant feel less like a form and more like a real helper. It’s a small touch, but for businesses, it means less user frustration and better conversion on...

Designing a product is one thing. Getting people to actually use it is another. We’ve focused on simplifying the setup so much that businesses can just scan a QR or paste a script — and their assistant is live and working. No UI walkthroughs. No complicated configuration. Just logic that adapts based on what the user writes. Still improving, but so far… less effort = more adoption. Would love...

Whenever a booking is made, our backend sends WhatsApp reminders at 3 days, 1 day, and 4 hours before the appointment. They’re only sent if: ✅ The business has connected WhatsApp by scanning a QR ✅ The customer shared their number voluntarily This flow runs separately from our n8n-based assistant, but it’s fully integrated into the experience. It’s a quiet feature — but one that reduces...

We’ve started building the assistant’s memory by storing key patient data — so it knows who it’s talking to, what’s already been asked, and what’s relevant to each interaction. This goes beyond just answering questions or booking appointments. The platform will also include an internal assistant for the clinic staff — capable of answering things like: 💬 “How many appointments are booked for...

The assistant now handles bookings on its own — asks the reason, finds the best doctor, and confirms based on real availability. To make sure people show up, a backend service also sends automated reminders 3 days, 1 day, and 4 hours before the appointment. This works through WhatsApp or a chat widget, and everything runs on a dynamic n8n workflow per business. The process feels instant, and...

This week I added a feature that lets each client load their own FAQs — and the assistant actually uses them to answer in context. No coding needed. Just question → answer → save. Internally, it turns into a reference the assistant pulls from when replying. The goal is to make it feel like it knows the business, instead of replying with generic fallback answers. Next steps: I’m planning to...

This week I managed to get something big working: you can now scan a QR code and instantly link your WhatsApp number to a fully functional assistant. The QR is generated using Evolution API, and once scanned, I dynamically create a workflow in n8n that handles incoming messages. Everything runs through WhatsApp, but I added a simple internal chat UI so I can test how the assistant behaves...

Many companies have chatbots, but almost none are truly connected to their information. We're developing a platform that allows you to have your own assistant, connected to WhatsApp or your website, and with access to your real information: schedules, FAQs, appointment calendar, etc. This week we're starting to see progress.