LettsTV

The brief

The editors needed a site that could publish quickly, look credible, and get their stories in front of people on Google and social media. They also needed it to keep working while the newsroom grew.

What I built

On the reader side there’s the home page, article pages, and topic pages that follow a story over time. Readers can sign in with an account they already have, and when a story comes with a report or a document, they can read the PDF right inside the article instead of being sent somewhere else.

Behind the scenes there’s an editorial admin where the team writes, edits and publishes without touching any code.

An AI newsroom, with editors in charge

News moves fast, and a small team can’t write everything by hand. So I built LettsTV an AI editorial pipeline. It does the repetitive work of preparing stories automatically, then hands each one to an editor in the admin. Nothing goes live until a person has read it and approved it.

The team gets the speed of automation and keeps full editorial control. That’s the balance I aim for whenever I put AI into a real business: it takes the busywork, and people keep the final say.

Getting found

For a news site, being found is everything. Every article goes out with the search and link-preview information it needs, so stories show up properly on Google and look right when someone shares them on WhatsApp, X or Facebook.

Running it

Being the only engineer on a live site means building new features while people are reading, fixing problems fast and keeping pages quick. I also moved LettsTV onto our own server with automatic deploys, as part of a move that cut hosting costs by 78%.

Tech

Next.js, React, Node.js, AI editorial pipeline, SEO & link previews, Social login, Docker, Nginx