SpokenLayer
Audio meets Visual. Building an AI-powered 3-product Suite for SpokenLayer, a full-service creative studio specializing in custom, short-form digital audio content.
MY ROLE:Ā Product Design and Product Strategy
PRODUCTS: B2B CMS, Internal CMS, AI Video-Generation Platform & End-User Mobile Experiences - (All 0-1)
JUNE 1, 2023: SpokenLayer Acquired Listory. š¤
Jump to Visual AIĀ POC >
THEĀ PROBLEM
For Publishers: Traditional news media no longer seems as relevant or dominant for the younger audience, as the demographic gravitates towards newer, "snackable" sources.
For Podcasters: Audio content discoverability is difficult, preventing podcasters from reaching & engaging new listeners. People are overwhelmed by where to start with podcasts.
THEĀ SOLUTION
We based our initial product offerings on one validated use case which was to create a platform that turns brand, publisher and podcaster long-form content into engaging, short-form (< 30 seconds, based on research of optimal lengths) video content, we call ārich media storiesā. They cater to a younger audience and can be easily distributable across various platforms (ie Tiktok, IG, FB, etc.) where this demographic hangs out.
THEĀ MARKETĀ GAP
No solution existsĀ that provides short-form videos with these key components in mind:
ā¢ Scalability
ā¢ Automation & Speed
ā¢ Optimized for High-frequency content pipeline
OBJECTIVES
- Build for scale and speed: leverage AI to automate steps in the workflow (summarization, narration and video generation)
- Validate end-user product experiences via implementation and measurement with our partnersĀ (brands/ publishers/ podcasts)
- Develop a client-facing CMS with a user-centered approach based on client usage and feedback.
- Design client-facing features to be accessible, intuitive and natural to their need and patterns
- Secure new monetization opportunities with AI-generated products (B2B)
- Goal for End-User Experiences:Ā Learn from engagement metrics on our experiences in order to iterate. We are currently in Q4 planning to discuss our KPIs (previously TSL- Time Spent Listening)
The Product Ecosystem
IĀ am currently building out 3 product entities in our ecosystem:
THEĀ INPUT The Client-Facing Project Management System (currently for client Intake & Delivery of assets). This is the latest addition to the product suite. We are working closely with our partners to iterate based on their feedback and workflow. Users can login/out, view project, check production status, sort, search, comment/view comments, and download deliverable (video asset), once available.
BRUNO
BEHIND-THE-SCENES The CMS & AI-powered video generation tool that allows: 1) viewing/selecting the long-form content, 2) summarizing it, 3)Ā narrating it, and 4Ā ) generating the short video. (the latter 3 all powered by AI).
THEĀ EXPERIENCES
THEĀ OUTPUT (please note: AI-generated audio removed from videos below) : The End User Experiences are our Video Products. They include 1) Animated Audio: Converting a script or audio into an animated waveform with static or lightly animated brand assets; 2)Ā Cinemagraphs:Ā long-form Script or audio turned into an animated loop using brand assets; 3) Ausums: Leverage our AI tech and professional human studio to create a premium videošš½.
TEMPLATE & OUTPUTĀ EXAMPLESĀ LEFT:Ā TechCrunch Scene types (ie. Title, Body, Display, Caption, Picture, List) and Template Design I created. Once this design recommendation was approved, I repurposed the designs for Ā implementation in our Bruno Studio as a TechCrunch template. RIGHT: Examples of still frame output from the video generator, for various articles and podcast episodes.
create business opportunity
šAI for Podcasts &Ā Video
Introducing Generative Visual AI
to SpokenLayer's capabilities
āTHE CHALLENGE:
āSpokenLayer's partners include a wide array of brands, publishers and podcasts. One such podcast is Shadow Realm, for which SL produced 2 seasons that brought the fantasy-fiction story to life. After podcast production was complete, and marketing was about to begin, it turned out Season 2's show art was nonexistent.
Around the same time,Ā our product team was experimenting and testing visual GenAI capabilities for integration into our video generation product, called Bruno. So we took on the challenge of bringing the show's visuals under our wing, and using it as an opportunity to test and validate a few assumptions:
āThe use of Visual GenAIā¦
- āEmploys our existing assets and repurposes them:Ā ie. (transcript > text prompt >Ā visual)
- Increases automation of our processes
- Saves time and costs for our client and our company
- Is scalable across properties
- Can be included as a service offering for clients
- Widens bandwidth of our content intake with clientsĀ
- Is integrable into our Video generation product.
THE OUTCOME:
š The POC was a success, checking off all our assumptions. I presented our process and accomplishments using AI capabilities at the following company all-hands meeting (Here are a few slides from that presentation), and the implications for our technology, product and company as a whole. The POC opened doors and since then, we have secured 2 new partnerships for the coming year. Season 2 launches Fall 2023 and includes all the assets we created including Episode Art.
