Skip to content

callme.lt

AI phone agent for businesses. My own product: an AI phone agent that answers a business's calls in Lithuanian, takes booking requests and hands the caller to a person when needed.

TechnologiesNext.js, React, TypeScript, AI APIs and automation, PostHog
callme.lt website screenshot

Challenge

Small businesses lose customers on calls nobody picks up: during a haircut, a dental procedure, a lunch break or after hours. A call centre costs too much for them, and few people call back after hearing voicemail.

Solution

I designed and built callme.lt from the first idea to a paid product. The agent runs on a Lithuanian phone number over SIP and answers only from the business's own information: hours, services, prices and FAQs. During onboarding a crawler reads the client's website and turns it into the agent's knowledge base. A NestJS backend handles agents, numbers, call history and Stripe billing, clients manage everything in a React dashboard, and the public site runs on Next.js. After each call the business gets a summary and a transcript, and the agent passes harder conversations to a staff member.

Results

A live SaaS product with public pricing: Starter from €99/month, Enterprise from €999/month.

One agent across phone, website chat, Messenger and Instagram.

Google Calendar and API integrations for bookings and lookups.

History, summary, transcript and a quality score for every conversation.

Worth knowing

Building my own product taught me something I now use on client work: the first setup decides whether anyone pays. The agent is only as good as the information it gets, so a large part of the work went into collecting that information from the client's website automatically.

callme.lt: screenshotcallme.lt: screenshotcallme.lt: screenshot
Next projectBalaretisJewellery online store for three markets

Have a similar task? See where the price starts.

Choose the services and see where the price starts. Send your enquiry in a minute, and I'll reply myself.