Hey channel insiders, I'm your host Katie Bavoso and welcome to this episode of Partner POV brought to you by Moments Lab. If you've ever spent hours digging through raw video footage just to find that one perfect clip, you're not alone. Whether you run a full-scale production studio or manage and create content for a media or sports organization or a brand, searching through content can be a huge time suck, especially when your archive isn't indexed properly.
But what if you could find exactly what you're looking for without even remembering the file name? That's where AI video discovery comes in. It's a practice that AI company Moments Lab is leading with its own indexing technology called MXT, which makes searching massive video libraries as intuitive as searching the web. And Moments Lab's latest innovation, a discovery agent, takes things a step further. Here to tell me more about this technology and how it saves time and money and even offers new sources of revenue is Phil Pettipal, CEO and co-founder of Moments Lab.
Welcome, Phil. >> Thank you. Hey, Katie. Thanks for having me. >> Thank you so much for being here. I hear you're recently a New York resident, so it's great to have you in this time zone. Have you tried the bagels yet? Have you tried the the pizza yet? It's important that we get this information from you early on. >> Absolutely. I'm performing my, you know, Starbucks skills order. It's really uh the one that is the hardest to get, I think. >> Yes.
Well, that's important. Now, you're a real New Yorker by ordering that on the mobile app. So you'll you'll understand the rush in the mornings trying to get that that order. Well, fabulous. I'm so excited to have you here today. I'd like to start with your company's origin story. Can you tell me what is Moments Lab exactly and when did you enter the market? >> Yeah, that that's that's a good question to explain what we're doing. So I originally came from the media entertainment industry and I used to work for u big TV networks and media networks as a broadcast engineer and my job was to make sure that journalists can work way faster with video and actually where you are really um in the core of building video you're not really well positioned to you know develop technologies on AI to make sure you can help.
So I took a step back. I quit my former company to make sure that I can create a team, team up with Fred, your co-founder and create something that will help all these content creators to work faster with video. Building video today is taking ages like it's taking probably 3 hours for a 3minut video and that's why video is a very complex signal. So very early in the beginning we understood that one of the big pain points that was blocking video from really being easy to work work with was indexing being sure that anyone can understand what's inside a video is today taking a long time for you and me understanding what's inside one hour video will take one hour but with AI it's changing totally the game because you can understand what's inside the video within maybe just one minute so it's opening new opportunities and one of the opportunities is For instance, searching for video and creating video, finding massive amount of content like you know this small three minute video you need in amount millions of hours of video and that's what we've been addressing for since the very very early beginning using our indexing technology. >> Wow.
So a lot of experience coming into this field knowing exactly what the pain points are. So who are your customers now? Can you tell me who the customers are that rely on these solutions from Moment's Lab? Yeah. So we're working with customers that are working with huge amount of video. So of course we're seeking first about media and entertainment companies. It could be tons Reuters Sinclair ITN in the UK here in France for instance. We're also working with sports right owners can be soccer club like bay leusen or any kind of sport like the world championship which is also one of our customer that we're helping them to find all the golden nuggets they might have is all their races but we're also working with marketing organizations like KPMG Amazon ads also that are part of our customer what they need to find very quickly anything about the specific product in massive amount of video >> exciting stuff so let's get to know the new technology on the block that you've been introducing.
Talk to me about AI video discovery. What is this and what need did you see in the market for AI technology that specializes in indexing videos? Essentially making videos searchable. >> Let's say you need to build a video of 3 minutes or let let's say 10-minute video. You know that 75% of the time you will spend to create this 10-minut video will be what we call sourcing content. You will need shots, lot of different shots. And actually in a 10-minute video now there's probably 100 different shots mostly to take these shots from video that already existing real video.
And some of our customer estimated that to find one shot it took them 5 minutes. So boom 100 shots time five and you have like 10 hours to just build a 10-minute video just for the sourcing part. Video discovery is a capacity to be able to understand video at scale. So that this step you need uh to build this video meaning sourcing content searching for content can be way faster and video discovery is a way to search among vast amount of video so that you can find exactly what you need and sometimes you are searching for something you don't really know.
If you take an example let's say you have 1 million hour video where do you start? Nothing is indexed. So you can watch video one by one or search through titles or names but most of the times there's nothing. So our job is to make sure that our users can find the content and that's what we call video discovery. So the ability to first describe index content and then being able to search for content. So that's two block we are proposing within our solution is the capacity to index and to search and to retrieve content. >> Very well put.
So listening to you, Phil, I imagine this is a great way to save time on all kinds of media projects. I myself was once an intern for a news organization in New York City many years ago, digging through news content, the actual plastic video tapes, trying to find B-roll of almonds for the 6 p.m. newscast. So I imagine that while this technology can save a lot of interns tears and bad days at work, I do also want to know what is the financial return on investment.
So the first thing we could think about is of course efficiency. You don't need to log content manually. So yeah, I can do that. That's working very well. It's probably five times faster than real time. You can do that do it really at scale with massive amount of video. So that's going faster. One of the thing that our customer are estimating where they're working with us is the time it requires to search and to find content. And that's probably the most interesting part here.
Well, the first one which is okay I can index automatically but the final outcome is more about how fast I will be able to retrieve content and then if we take like a step back on a higher level most of our customer are repurposing content so you have 1 million that of content 1,000 hour content you're a brand you need to repurpose on to take all the masters from a campaign for a brand or product for instance and you want to create from the master video like 10 different videos you're going to repurpose on social media.
And that's something that you can do now at scale with video discovery because you can really nearly just pro what you're looking for. Let's say I want to build um sub story about this main story just three minutes and you will be able to repurpose it very very fast very quickly. That was not really possible before because you needed lot of people to be able to do it and I'm not sure that AI would have matched business reality here. So the capacity of being able to do that at scale is really what is the higher ROI that we can expect here is this capacity to monetize videos that already existing into different formats to social media. >> So Phil, I'd like to talk about the kinds of businesses you see benefiting the most from MXT.
What are those businesses? >> These businesses are the one that are working with a lot of video. Uh can be creative organization, brand content organization, media rights organizations. uh and they can be spread across lot of different industries. It can be to media, entertainment, sports, brands, enterprise communication and most of the time all these organizations they need to find they need to search content they need to repurpose content and when you need to find and you need to repurpose content that mean that you need video discovery.
So you need an AI indexing like MXT to be sure that your video will be findable and searchable forever. >> So let's bring this into the real world. I'd really love to hear a success story in which Moments Lab was able to help a real customer with this technology. Can you think of anything for an example? >> Yeah, we I actually we probably could take two different examples. So the first one is maybe media entertainment. So I will mention a couple of them that exactly share the same success story.
So it can be Sinclair, it can be Thompson Reuters for instance or ITN and where they're coming to us. It's it's very similar use cases. They have massive amount of content. sometime even plastic tapes you mentioned before and they want to digitize all of that and it can be sometime 50 years 100 years of archive our biggest customer have like 100 years of video archive so it's huge it's massive and especially in the US um in media entertainment there's a lot of station TV station in cities and town and no one knows exactly what's inside and so the goal here is to give a second life to content to make sure that among these pabytes of hours of video I can prompt find something that is relevant tell a story very quickly and repurpose all of that and that's it's super emotional because we're able to find back content that everyone forget about and uh that's that's super super powerful.
So that's that's part of um more media entertainment use cases and the second one is is more about brands and marketing organization where actually the most important thing for a brand is is is is its product and they want to be able to find back in all amount of video they have whether it's 1,000 or 1 million every time that I've seen this Nike Air Max 1000 I want to extract that to build another video and this is taking a very very long time And sometimes we've been discussing with brand that doesn't even know how to search and they're saying well actually I need to pay the service a service company to make sure I can find back my video and paying humans to log it manually.
So this is some use case that we've been working on recently and uh on making a world video archive searchable is very very rewarding for the customer but also for us because we know that we can help them building video faster and building video they were not able to build before. How long does it take to get up and running with Moments Lab's AI video discovery platform and what can customers and potential customers that is do ahead of time to make this process easier? >> Yeah, so it's a three-step process actually.
So the first thing we we are doing um is what we call the scanning step. We scan all the inventory of archive and try to list everywhere there's archive. Sometimes you know it's on hard drive, sometimes it's on tapes and we kind of centralize everything and this step can be pretty quickly if everything is listed correctly. Uh and that start to have a pool a bulk pool of ours that we can we can start to index. The indexing part is very quickly and we also most of the time customer are thinking about the migration process taking long technical steps.
Actually the longer step and the harder step is not really about technical indexing. It's more about setting the right change management policy in place and to do that uh what we like to do is to involve people that are going to be the future users. We are bringing AI as another member of their team. So okay, welcome MXT. This is our AI indexing technology as a new member of your team. Don't consider it as a technology that is going to really replace you, but someone that is going to help you going faster and someone that is going to help you finding faster content and the way that we are on boarding AI um is working very well and that's going way faster into tent management because what you don't want in tent management is people that are not being involved into the process that can have defiance or that might not be very happy being seeing AI in incoming into their businesses without them.
So that's really a crucial step that needs to be managed correctly and there's a good playbook about that. The last step and it's also helping step number two about tend management is being able to define clear success metrics where are we expecting the AI to generate revenue to save time and this should be said nearly at the be than at the beginning when we are qualifying an opportunity that's probably one of the first thing we are trying to to set here it's not about magic it's about really delivering success and okay the world AI feels magic but actually we need to be sure that we can drive uh revenue venue.
We can drive um savings here and that's a very very very important step. >> Let's talk more about the moments lab discovery agent. What problem does it solve for teams working with video? >> Yeah, so video teams are looking to either extract insight or intelligence from their video or to build what we call the rough cut and building the rough cut. The rough cut is nearly the first version of the video. It's a raw edit, but actually it's nearly 70% 80% of the video.
And to do that, well, you need to find all the shots you need and you need lot of people to work on to find content. And so the discovery agent discovering all of that. So it pinpoints every moment, extract every moment and put that into a timeline. That's what we call the rough cut. Our goal here is to make sure that the time between the very moment that you are starting to write a brief and you have a rough cut is as minimum as possible as fast as possible to achieve to the very first moment that any editor will hit the space bar and play on this rough cut is something that we are committed to work on to make sure that building a video can be as quick as possible. >> Phil, let's get a few more details about this discovery agent. how would I as a consumer use it?
How do I interact with it? >> So, it's it's very simple. Uh it's very human. It's like talking to a colleague where you say, "Well, I'm building a video about Master Chef um and I need to build a highlight about the last season. Find me the best highlights on anything that can be interesting for a specific audience." And then well the agent will find all these shots and maybe even sometimes suggest other shot that might be interesting for you and compile that into a rough cut.
So very easy, very natural. It's like talking to a colleague. >> Phil, in thinking about that Master Chef example that you just gave where you can build out a highlight reel, give me an example of that because immediately I'm thinking bloopers. Could I realistically ask the agent to find all the mistakes from that season and build that into a sequence of video? >> Yeah, absolutely. So, the agent is is have a perfect understanding of what's inside the video and what it's all the season of of the show for instance.
But also the agent has the capacity to crawl the web. So if we go back to this this Master Chef example, let's say there's a lot of comments about a specific moment in the last season of Master Chef that have been commented a lot into social media. Well, the agent has the capacity to crawl through the external data source that are available on the web to cross reference what have been shared by fans on what's inside the video library so that we can really do the perfect matching on having very unique moment that have been told and discussed a lot on the web. for instance. >> I love that.
I would immediately want to work on that project knowing that I could have the help putting that together. An excellent example, Phil. Phil, let's talk about your relationship with Amazon Web Services. How does AWS support you at Moments Lab as a channel partner and how does it help you stand out in the market? >> We have a very deep connection with AWS um on the partner side, on the technical side and many many fun levels. on the channel partner side.
Well, the first thing is that we've been helped a lot um to connect to the industry to connect to key players in industry as a trusted partner and this is super powerful because that's saving time and enable really um trusted connection with the customer. The second thing is helping to evangelize AI. Agentic AI requires a lot of change management and having AWS being able to help us on customers to evangelize the market to share best practices is something that can also bring back trust into AI and agentic AI and that's something that is super powerful and the last thing is really probably around build helping us building our awareness across the industry. uh we have uh one of the best AI indexing for video uh technology on the market and AWS is really really helping a lot to promote our technology across the market here. >> Fantastic.
And going on from there, speaking of channel partners, Phil, what kinds of channel partners is Moment's Lab looking for to help distribute your solutions? >> Yeah, so we're working a lot with channel partners uh and there's three kind of type of partners we we are working with. The first one are tech partners that can integrate our technology into an OEM mode so that it's it's very seamless and they're using our API. The second thing are um system integrators. um so system integrator that are integrating a video discovery engine among other part of the workflow with a storage system uh with I don't know um any uh information system that can help to drive insight for instance and it's kind of part of a global solution and also uh consultancies agencies and advisor that can help I would say customer into their different technology decision and that's kind of the three level of partnership we can have today on the market. >> Well, I know that I've got a billion more questions for you about MomentsLab and the technology that you're bringing to market, but if I'm an interested viewer, where can I go to learn more about MomentsLab and get started? >> Yeah, please go on momentslab.com.
Uh there's lot of sources, lots of resources that can be helpful here and please don't hesitate to reach for any demos. We'll be very happy to help. >> Phil, thank you so much for your time today. It's been a pleasure speaking with you and I'm so excited to follow along with Moments Lab through the rest of the year and beyond. >> Thank you, Katie. >> And thank you at home for joining us. To see more episodes of PartnerPls.com. Once again, I'm Katie Bavoso and I'll see you next time.
This transcript was generated automatically from the
video's captions and may contain errors.
