Newsly

The problem

Currently, many companies, managers, politicians, and individuals who work and need to consume a lot of news to make decisions or take action. A significant percentage rely on press clipping companies or media monitoring agencies, but these companies are already falling behind, losing out to AI and social media. At the same time, users who no longer use these services spend more time consuming, receiving, and filtering the necessary news (for example, politicians who are unaware of a particular event from the previous day).

Therefore, there is still a disconnect between the speed of news and the internet, and AI filters to give users a better and faster filter on the most relevant news they need at any given time.

The global media monitoring and clipping market is valued at approximately USD 5.99 billion in 2026.

The solution

Newsly is a web platform that aims to bring news clipping services to the pocket of every person and company without the need to contract services for extended periods, and at the same time, receive personalized news according to immediate interest.

To achieve this, the platform allows users to search for news in two ways: by specific topic (ideal for users who need access to specific news) and via AI search (for people who want a summary from various reliable sources).

After searching for and selecting the news, the platform generates a summary using AI, extracting and displaying the most relevant information so the user can understand the news as quickly as possible. If the user is engaged in another task, they can listen to the summary (also powered by AI), facilitating comprehension without distracting them from their current activity.

How was it developed

I used the MeDo platform for the entire system, following specific instructions. I gave it a clear context: I was building a news clipping web service and needed a news search engine, a summary for each news item (otherwise, it would just be another search engine), and the option to listen to the summary (very relevant for busy users). The platform first generated a detailed list of requirements, which I accepted, and then created the app.

Next, I added the search engine using AI (AI search skill), and finally, I asked it to run a full test (it detected some JavaScript issues when generating the summary, corrected them, and everything worked).

The use of MeDo

Initially, I selected the news skills to retrieve news by topic, LLM to create summaries, and text-to-speech to read the summaries aloud.

But the most interesting skill I worked with was AI search, as it allowed me to search for news more quickly and from different sources in one place, making it quite practical for very specific news stories that require more context.

Another very practical feature for me was the testing, since the tests (unlike other Vibe Coding platforms) simulate the complete action of the users, and traceability can be done (including screenshots) using the AI Quality Analysis feature.

Built With

• llm
• medo