Introducing Lucia

Everything you need to know about Lucia, BLAM’s new AI assistant.

The Blue Lucy team introduced BLAM’s new AI assistant, Lucia to the media and entertainment industry at IBC Show 2024 in Amsterdam. And, while the application is still in development, the industry response to Lucia has been overwhelmingly enthusiastic, even earning the Blue Lucy team a TVBEurope Best of Show IBC 2024 Award. But, in a market flooded with AI messaging, what makes Lucia special and what difference will she make to BLAM users? We asked Blue Lucy’s head of product, Joshua Martin and Lucia’s primary developer, Aaron Leanage, to answer some frequently asked questions about Blue Lucy’s newest addition.

Why did you create Lucia?

At Blue Lucy we believe that technology needs to constantly evolve to keep up with the needs of the media and entertainment industry. This ethos is a big part of the reason why we built our BLAM platform on a microservice architecture, as it allows our team to develop and deliver new integrations and capabilities really quickly. But we wanted to explore how the latest AI developments could make it even easier for our customers to get value from the platform.

Our goal was to lower the entry point to automation – so that customers don’t always need an engineer to create or manage processes or get answers to complex questions about workflows. Most operators can describe what they’d like the system to do, but not everyone has the time or capacity to learn how to access BLAM’s full functionality – Lucia was created to remove this barrier for our less technical users. Even typing is optional because Lucia responds to both text and voice prompts in conversational language, making BLAM accessible to almost anyone.

What can BLAM users expect from Lucia?

Lucia works alongside organisations’ predefined workflows to help with anything from simple searches to bulk actions and complex chained commands. Where workflows in BLAM are hard-coded to do things in a pre-defined way, Lucia provides flexibility for use cases that change from day-to-day.

For example: if you’ve got a big batch of new content coming in and you need to generate placeholders, remove the bars and line-up and then save the media in a new storage folder, Lucia’s got it covered. And she remembers what you’ve asked her to do before, so it’s really easy to repeat processes or revise them without having to start from scratch.

Some of the actions Lucia can take (which can be combined to create complex chained commands) include: asset searches, folder searches, creating placeholder assets, creating folders, attaching and detaching assets from workspaces or folders and getting metadata, workflow or user data.

Besides automation, what other capabilities will Lucia provide?

Because Lucia is based on a large language model (LLM) she can be used to perform tasks like summarising information, doing mathematical calculations, translating text and answering general questions that users may have about content formats, metadata and almost anything else – and you don’t have to leave the platform to find the information you’re looking for. Lucia can also be used as a business intelligence and support tool. You can ask her questions about your content and workflows, like how many assets are in a certain folder or whether any episode numbers are missing from a series of programmes, or get Lucia to perform calculations on data retrieved such as “get me all complete workflow runs completed in the last 24 hours and calculate the average runtime.”  This kind of information would previously only be available if the workflow had included a report generation process. 

As Lucia is still in development, we’re still in the exploratory phase, establishing what the potential capabilities are and what would offer the most value to our customers. For example, if the AI assistant is integrated with BLAM’s data processing service, Lucia can use historic data to predict asset-based costs and identify workflow bottlenecks. Similarly, Lucia could rewrite BLAM user documentation in everyday language tailored to your operational context and answer users’ technical queries, cutting down on customers’ first-line support costs.

Have there been any unexpected benefits?

It quickly became apparent at IBC Show that different BLAM users would find value in Lucia for different tasks. So, while a user may be interested in asking her how to do something or creating bulk or chained operations, managers may find Lucia’s ability to summarise large amounts of data more useful and use her to find out how long it takes for workflows to be completed or how long it may take for a user to pick up a task. One of the unexpected benefits we’ve discovered is how Lucia could be used to optimise BLAM’s implementation, for example an administrator can use Lucia to identify which microservices need to be updated or to interrogate workflows to identify bottlenecks.   

Is Lucia safe?

Lucia can only add value to content or processes in BLAM. She can’t modify workflows or change the original assets, and she isn’t able to do anything that results in a loss of data. What this means is that, while Lucia can be used to create new versions of content, the original asset will remain intact and unchanged and, although the application can be used to build complex chained commands, she can’t interrupt or edit existing workflows. Lucia is also bound by the same user-permissions as the operator working in BLAM, so she can’t take any actions or access any information that the user isn’t already cleared for.

When it come to keeping your data and content secure, Lucia uses your organisation’s AI provider, credentials and authentication so there can be no cross-contamination of data between BLAM customers, and Lucia will comply with any security agreements you have in place with your provider. The application currently runs in OpenAI’s ChatGPT and can be set up in Microsoft Azure’s Open Ai, but we plan to integrate Lucia with our customers’ choice of any large language model (LLM.)  

When will Lucia be available in BLAM?

Lucia is scheduled for general release to BLAM customers by the end of 2024.  Get in touch now to chat with the team about how Lucia might benefit your organisation or to arrange a demo.