Vibe coding with Lovable
Thomas Vervik, Software Engineer at SeniorDev
Build AI agents with Model Context Protocol
Hennadii Tarasenko, Sr. Software Engineer at SeniorDev
Automation with n8n
Iryna Mohylenko, Product Designer at SeniorDev
Build AI agents with MCP and Vibe coding with Lovable
Navigating the Future: AI Agents and the Model Context Protocol
Thomas Vervik er daglig leder i SeniorDev, og erfaren utvikler. ��Han har sammen med kunder bygget et utall applikasjoner og plattformer, og jobber aktivt disse dager med å teste ut KI verktøy for å finne ut hvordan de beste skal brukes for å bygge suksessfulle IT verktøy fremover.
Thomas Vervik
Navigating the Future: AI Agents and the Model Context Protocol
Prompt-Based Development
About Lovable
Navigating the Future: AI Agents and the Model Context Protocol
Requirements�I have an existing Webflow site with events, stored in Webflow’s headless CMS collection
What we are going to build
7
A custom event registration system
Navigating the Future: AI Agents and the Model Context Protocol
I would like to build a simple event system where users can register to an event.
The event organiser will be "SeniorDev Knowledge Corner"
The system should show a list of events.
Each event should have:
- Title
- Description
- Topic
- Date and time (a string)
- Event image
- Link to google meet
Show the events as a list, not as a grid
Initial Prompt
Navigating the Future: AI Agents and the Model Context Protocol
…
Connect to Supabase
Navigating the Future: AI Agents and the Model Context Protocol
We want to have control on the schema and tables created
And also to be able to check the data
Connect to the database
Navigating the Future: AI Agents and the Model Context Protocol
Use the theme in the attached file
Style the application
Navigating the Future: AI Agents and the Model Context Protocol
We want to fetch the events from Webflow's headless CMS collection.
The URL to use is: https://api.webflow.com/v2/collections/:collectionId/items
Store it in our database. The mapping from the response to our table should be:
fieldData.name -> events.title
fieldData.description2 -> events.description
fieldData.speakers -> events.topic
fieldData.cover-image-2.url -> events.image
fieldData.date -> events.human_date (the column type should be text)
fieldData.event-date -> events.date
fieldData.link-to-event -> events.link-to-presentation
fieldData.google-meet-link -> events.google-meet-link���Add the collection id to the edge function instead of taking it in as a parameter��I also want to set the collection id as a secret��Add first the collection id to the edge function
I don't see it in the edge function, where is it?
Great! Then I want a synchronize button to fetch the events in the collection
Connect to Webflow CMS Collection
Navigating the Future: AI Agents and the Model Context Protocol
Don't HTML espace the description and topic, show it as is��Organise the event card as on the screen shot
Add the event image under the title. Don't cut the height
The speakers in div on line src/components/EventCard.tsx:78:8, they are hard coded, just remove them��
List the newest events first. Use the column events.date for that
Don't show the "Join Google Meet" button of no link is available��Move the topic section below the description
Style the event card as attached file
Navigating the Future: AI Agents and the Model Context Protocol
We want users to be able to register to the event. All we need is their email. And store it in the database, in a table named "registrations"
Show all event registrations on a new page.
Store registrations
Navigating the Future: AI Agents and the Model Context Protocol
Final thoughts
7
A custom event registration system
Navigating the Future: AI Agents and the Model Context Protocol
Hennadii Tarasenko er full stack-utvikler med over 15 års erfaring. Han spesialiserer seg på applikasjonsutvikling, AI-integrasjon og optimalisering av systemer med høy trafikk.
Hennadii Tarasenko
03
02
01
Navigating the Future: AI Agents and the Model Context Protocol
MCP is used to develop AI agents that can interact with proprietary datasets seamlessly.
It is applicable across various industries, enabling the creation of smarter and more efficient automation tools.
Practical Applications
MCP provides a standardized framework for feeding AI agents with custom data.
This allows AI systems to provide responses that are tailored to individual user needs and contexts.
Key Features
Purpose of MCP
The Model Context Protocol (MCP) enables AI agents to utilize structured data effectively.
Its goal is to replace traditional web interfaces with intelligent, data-driven interactions.
Introduction to Model Context Protocol
5
Navigating the Future: AI Agents and the Model Context Protocol
Will create a system that summarizes the essence of a youtube training video
The MCP client is Claude. MCP server is created in Python
workflow with MCP tools:
Today we will build a demo MCP server
Navigating the Future: AI Agents and the Model Context Protocol
Iryna Mohylenko er produktdesigner i SeniorDev, og har det siste året aktivt utforsket og brukt KI verktøy for design og nettsider.
Iryna Mohylenko
Navigating the Future: AI Agents and the Model Context Protocol
1. Automated flow: Post both text and an AI-generated image on Facebook
Automation with n8n
7
Navigating the Future: AI Agents and the Model Context Protocol
2. Automated flow of replying on support emails
Automation with n8n
7
Navigating the Future: AI Agents and the Model Context Protocol
n8n makes it easy to build flexible, no-code automations that connect tools, process information, and save time. Whether it's content publishing or customer support, you can combine AI and integrations to create smart, scalable workflows without writing full code. It’s a powerful way to reduce manual work and focus on what really matters.
Final thoughts - Automation with n8n
7
Navigating the Future: AI Agents and the Model Context Protocol
Sr. Software Engineer at SeniorDev
Hennadii Tarasenko
CEO at SeniorDev
Thomas Vervik
Thank you!
9
Product designer at SeniorDev
Iryna Mohylenko