SymphonyAI
Palo Alto-based SymphonyAI is leading the way in AI data collection and analysis, and its market insights could have wide-reaching implications.
Launched in 2017, SymphonyAI is a global enterprise AI leader, with 3,000 employees in more than 30 countries. Active across five industry verticals including media, fintech and retail, it builds AI SaaS (software as a service) products targeted to solve specific business problems.
Explaining what this all means for the C21 community, Marc Liebmann, chief revenue officer, media, at SymphonyAI says it’s about “helping media companies deploy their content more effectively. We help creators and IP owners distribute their content better across platforms and channels. We make it possible to optimise what they own, and gain a clearer understanding of their content’s engagement with audiences.”
SymphonyAI’s bespoke media platform Revedia is evolving with the advancements in generative and predictive AI (of which more later), enhancing its core data collection and analysis capabilities, says Liebmann. “The job starts with unifying a company’s data, aggregating and normalising all first- and third-party data into a single source of truth. This is not a new problem, but it’s difficult in media because there is such a lot of messy data. And messy data is useless data, because you can’t trust it.”
This data cleansing process leans heavily on SymphonyAI’s proprietary tools, explains Liebmann. “Historically, all of this has been done manually, and that means you get a lot of errors such as mismatched titles. It could be something as simple as the first run of a show being described variously as ‘season one’ or ‘series one’ or ‘S1’. AI deals with that.”
The next critical phase after having organised the data is delivering it in a way that is useful to all end users within an enterprise. “You need to put the data into the hands of the people who need it,” says Liebmann, “in a way that makes it easy to generate insights without having to go to an analyst or a data engineer to get help. We have several ways people can do that, such as drag and drop dashboards, automated report generation or our Media Copilot, an AI chat assistant that lets users interact directly with the data.”
As a topline illustration, Revedia could help a distribution exec stay on top of developments in different regions: “I log on in the morning,” says Liebmann, “I see my dashboard, and it shows me my KPIs and how my revenue is doing. I can then ask the AI copilot to compare the performance of my FAST channel in the UK with my FAST channel in the US. If there’s a revenue drop in the US, I can ask what happened.”
Liebmann says SymphonyAI typically categorises the capabilities of its system in three stages: manage, optimise and predict. “Management is about minding the store. It’s being able to understand KPIs, ensure compliance, all the basic things you need to run your business. Optimisation is where companies have historic data and want to use that to improve upon their KPIs. Here, the platform can help them understand what content they should be licensing, which distribution channels perform better for them, or even how better to negotiate the terms of their license agreements.”
Prediction is where things get really interesting. “This is where the Revedia platform helps with forecasting,” says Liebmann. “It can make predictions that go beyond just the historical data. There is another stage after this where gen AI is going to unlock new levels of personalisation and automation, for example the automatic generation of FAST channels with personalised schedules. We’re not there yet, but it’s coming.”
Could Revedia help a distributor with sales projections? For example, if a third-party producer is seeking deficit financing, can it help a company decide if that is viable? “It can,” says Liebmann, “assuming the project is aligned to what the distributor has been doing historically. We’re not just jumping into a wholly different framework. Users of the platform can look back at content that has done well for them in the past and understand what they should be paying for it and how much they might make from it.”
Most of SymphonyAI’s clients are confidential, but one that Liebmann can talk about as a point of reference is US media giant Hearst Television: “We started working with Hearst on linear TV channels and have migrated with them to digital. And now we manage the company’s FAST channel data. We do all the revenue management around making sure they’re getting paid correctly, as well as the data normalisation and performance insights.”
The Hearst example shows that Revedia can operate at scale, but Liebmann stresses that the client base is varied in size. “We have many small to medium clients as well as large global media organizations. This is a solution that really supports companies of any size. In fact, we sometimes find that it is the smaller companies that have more of a need to sort out their data.”
Liebmann says the need for an AI-powered solution like Revedia has grown with the fragmentation of the distribution landscape. A lot of the company’s work in recent times, for example, has involved FAST, as companies seek to achieve a competitive edge in a tight margin business. The core of the job is around improving channel performance, but Liebmann says the platform can help users formulate new channels: “It’s not a service like Amagi or Wurl that will actually generate a FAST channel. But Revedia can help you understand what’s the most likely concept to perform well.”
The platform can also help clients make sense of social media platforms like YouTube, at a time when more TV companies are shifting content to that platform: “Essentially, we ingest data from any source. That’s important because some platforms give very little data, and others give a lot more. What all this data allows us to do is generate directional insights. These are not exact, because we’re extrapolating data. But it means users can make rational decisions about platforms that are not open with data.”
SymphonyAI is in a position of strength, says Liebmann, because of the depth of its engineering expertise. It has a centralised team working on the company’s award-winning core AI platform Eureka, and then dedicated teams developing customised solutions such as Revedia for each vertical. In the media space, “we’re definitely heading towards more automation and predictive analytics – which is something that will be more apparent in 12 months. Before then, we are very excited by a major platform upgrade which is launching in Q1 2025.
“This will unify all the different types of distribution, so users can really understand the value of an asset. It will deliver a whole new user experience built on our learning over the last few years. We’re also expanding our Media Copilot to generate data visualisations. That will save a lot of time for people who need to put together a presentation for their management team. And we’re continuing to improve the AI title matching algorithm. That sounds very ‘meat and potatoes’, but data normalisation is one of the biggest problems companies have right now.”
The impact of AI on media comes in many forms, of which SymphonyAI represents just one, stresses Liebmann. “There’s a lot of buzz about the use of AI in the new Tom Hanks movie. That’s not what we do. But what I’m seeing is that media companies have got a lot more engaged over the last six months. Before that, they were testing the water. Now they are saying: ‘We’re in this for the long haul. We need a solution’”.
