Technology doesn’t stand still, and neither do we.

To make the most of Large Language Models (LLMs) like ChatGPT and Gemini, we need to ensure that they have access to the most up-to-date information and can interact with external systems, transforming them from simple text-generation engines into powerful and flexible AI Agents.

It seems I've developed a bit of a habit of getting a new CMS and immediately having to build something with it. The CMS that has our attention at Bitovi lately is Sitefinity, which is a bit of a departure from our usual CMS experiences. From a cloud-hosted demo instance of Sitefinity, content editors and page designers can get comfortable with the admin UI and start building pages. This post will introduce you to the main features for content in Sitefinity and also get you started with running the app and admin interface locally, so you or your dev team can get started on making widgets.

I recently had the chance to build a moderately complex React form. Instead of scattering the form logic across multiple components, I consolidated it in a single shared store using useReducer.

Understanding a large codebase is challenging, especially for new engineers. Traditional documentation often requires deep product and implementation knowledge, making onboarding slow and frustrating. AI-powered tools now offer automated solutions that generate documentation directly from source code, aiming to bridge this gap.

When a Fortune 500 financial services company approached Bitovi to modernize its design system, the team was drowning in prototype complexity. Designers were spending 50% of their time building prototypes for every interaction, state change, and validation message. The culprit? A design tool that forced everything to be a functioning prototype.

Building AI agents can quickly become complex and confusing. At Bitovi, we're always looking for ways to streamline efforts for all of your production use cases, not just that first POC. That's why we've developed techniques to help ensure your agents don't get confused as you scale and give them more to do.

Trying to finish personal projects can be tough. Seeing a graveyard of half-finished repos on your GitHub is discouraging. Often, the problem starts with jumping into a project without planning. No clear plan, just a repo and a dream. Or maybe you start with a plan but struggle to track progress. Regardless of why you're struggling, a bit of project management can really help.

Google’s newest AI tool, Stitch, promises to speed up the product design process. But does it actually accelerate design, or does it create “design debt” that organizations must reckon with later in their product lifecycle?