WaffleVideo.Ai
A marketplace connecting AI companies with creators for purpose-built, rights-cleared video training data.
Website: https://wafflevideo.ai/
Cover Block
Open sources
| Attribute | Value |
|---|---|
| Company Name | WaffleVideo.Ai |
| Tagline | A marketplace connecting AI companies with creators for purpose-built, rights-cleared video training data. |
| Headquarters | Los Angeles, United States |
| Founded | 2024 |
| Stage | Pre-Seed |
| Business Model | Marketplace |
| Industry | Deeptech |
| Technology | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (3+) |
| Funding Label | Undisclosed (total disclosed ~$250,000) |
Links
Open sources
- Website: https://wafflevideo.ai/
- LinkedIn: https://www.linkedin.com/in/joseph-newfield/
- Instagram: https://www.instagram.com/wafflevideo.ai/
- Reddit: https://www.reddit.com/user/WaffleVideo/
What an Investor Needs First
Open sources WaffleVideo.Ai is an early-stage marketplace attempting to organize the chaotic supply of video data for AI training, a problem that becomes more acute as models shift from text to multimodal understanding [wafflevideo.ai]. Founded in 2024, the company connects AI developers with a distributed network of creators who film specific, rights-cleared video clips through a mobile app, aiming to provide a more scalable and legally secure alternative to web-scraped or stock footage [Perplexity Sonar Pro Brief]. The founding team, led by CEO Joey Newfield and COO Joshua Mesnik, brings backgrounds in filmmaking and creative entrepreneurship, with technical leadership from Founding CTO Mario Aburto [postPerspective, 2026]. The company has participated in the FoundersBoost accelerator and lists Hallstone Ventures as an investor, with a pre-seed round estimated at $250,000, though full capitalization details are not publicly disclosed [ProjectStartups, 2026]. Its business model hinges on taking a marketplace fee from AI companies while sharing licensing revenue with creators, who reportedly receive a 30% share [Perplexity Sonar Pro Brief]. Over the next 12-18 months, the key watchpoints are the transition from a beta app to a public launch, the signing of disclosed anchor customers, and the validation of its patent-pending workflow claims against established data vendors. Partially corroborated -- Core product description and team are confirmed; funding amount and key commercial terms are from single, unverified sources.
Taxonomy Snapshot
| Axis | Classification |
|---|---|
| Stage | Pre-Seed |
| Business Model | Marketplace |
| Industry / Vertical | Deeptech |
| Technology Type | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (3+) |
| Funding | Undisclosed (total disclosed ~$250,000) |
Inside the Company
Open sources
WaffleVideo.Ai was founded in 2024 and is headquartered in Los Angeles, California. The company operates as a marketplace, connecting video creators with AI companies that require rights-cleared footage for model training. According to its website, the founding mission is to mobilize a global creator network to produce "commercially safe, purpose-built footage" for AI development [wafflevideo.ai].
The founding team consists of three co-founders. Joey Newfield, the CEO, is identified as a filmmaker and technology enthusiast. Joshua Mesnik serves as the COO, with a background spanning technology and entertainment. Mario Aburto joined as the Founding CTO, leading the engineering team [postPerspective, 2026]. The company's early development was supported by participation in the FoundersBoost accelerator program in Fall 2025 [ProjectStartups, 2026].
Key operational milestones are centered on product and team development. By January 2026, the company had publicly listed its expanded founding team, including Aburto as CTO [Perplexity Sonar Pro Brief, Jan 2026]. A beta version of its mobile application was available for creator onboarding and mission completion as of February 2026, though it was not yet listed on public app stores at that time [Quinnipiac University, 2026].
Partially corroborated -- Company details confirmed via its website and founder profiles; accelerator participation and team expansion are cited in secondary directories. The legal entity structure and specific incorporation date are not publicly available.
Under the Hood
Reported and inferred
The core proposition is a two-sided marketplace, but its operational mechanics are defined by a mobile-first creator workflow. Waffle Video connects AI companies needing specific video training data with a distributed network of creators who film it, using a proprietary app to assign tasks, collect footage, and manage payments [wafflevideo.ai]. The company's public materials consistently frame this as a solution for 'purpose-built, rights-cleared' visual data, a direct response to the licensing and specificity challenges in current AI training sets [Perplexity Sonar Pro Brief].
The creator journey, as detailed on the company site and in recruitment posts, follows four steps. Creators first onboard as 'WaffleMakers' and agree to terms. They then film assigned 'Missions' through the mobile app, upload and tag the clips, and receive payment for approved submissions [wafflevideo.ai]. A key economic incentive is the promise of ongoing revenue: creators are told they will receive a 30% share of licensing revenue, including from relicensing, which positions the activity as a potential source of passive income [Perplexity Sonar Pro Brief]. The company states its beta mobile app was available for use as of February 2026, though it was not yet listed on public app stores at that time [Quinnipiac University, 2026].
Technological differentiation is claimed but not independently verified. The company describes its mobile app and Missions system as 'patent-pending,' though no corresponding public patent filings were located in the available research [Perplexity Sonar Pro Brief]. The tech stack can be inferred as mobile-centric, requiring secure upload and tagging functionalities, but specific platforms or underlying AI tools for video validation are not disclosed. The absence of named enterprise customers or public case studies makes it difficult to assess the real-world throughput, data quality controls, or integration capabilities of the platform beyond its described workflow.
Partially corroborated -- Product workflow and creator economics are described consistently across the company's controlled channels, but key technical and commercial claims lack independent corroboration.
Market Research
Open sources
The market for purpose-built video training data is emerging as a critical bottleneck for AI companies developing visual models, creating a direct commercial opportunity for structured data marketplaces. This demand is driven by the shift from training models on vast, uncurated internet scrapes to using high-quality, rights-cleared footage that is commercially safe and tailored to specific use cases.
Third-party market sizing for this niche is not publicly available. However, the broader AI training data market, which includes text, image, and audio, provides a relevant analog. According to a 2025 report from Grand View Research, the global data collection and labeling market was valued at $2.22 billion in 2024 and is projected to grow at a compound annual growth rate of 25.7% through 2030 [Grand View Research, 2025]. The video segment is a subset of this, with its growth tied directly to the proliferation of multimodal AI models.
Demand is anchored by several clear tailwinds. The primary driver is the rapid advancement of generative video models from companies like OpenAI (Sora), Runway, and Pika Labs, which require massive, high-fidelity video datasets for training [The Information, 2024]. Concurrently, increasing legal scrutiny over copyright and data provenance, highlighted by lawsuits against major AI developers, is pushing companies to seek ethically sourced, licensed training material [Reuters, 2024]. This creates a wedge for marketplaces that can guarantee rights clearance, a point WaffleVideo.Ai emphasizes in its marketing [wafflevideo.ai].
Adjacent and substitute markets include general stock video libraries (e.g., Shutterstock, Getty Images) and synthetic data generation platforms. While stock libraries offer existing footage, they often lack the specific, mundane scenes needed for AI training and do not provide a workflow for custom creation. Synthetic data platforms generate video algorithmically, which can be cost-effective for certain scenarios but may lack the real-world variation and authenticity required for robust model generalization. The regulatory landscape remains fluid, with ongoing debates in the U.S. and EU about AI model transparency and data usage rights, which could further incentivize the use of auditable, licensed data sources.
| Metric | Value |
|---|---|
| AI Training Data Market (2024) | 2.22 $B |
| Projected CAGR (2024-2030) | 25.7 % |
The chart illustrates the significant growth trajectory of the foundational market, though the specific video data segment remains uncaptured. The high growth rate signals investor and corporate appetite for solutions that address the data supply chain, validating the core problem WaffleVideo.Ai aims to solve.
Partially corroborated -- Market sizing is inferred from an analogous, broader sector report. Tailwinds are cited from independent news coverage of the AI industry.
Competition and Substitutes
Reported and inferred
WaffleVideo.Ai enters a market where the primary competition is not a single direct rival but a fragmented set of alternatives, each solving a different piece of the video training data problem.
| Company | Positioning | Stage / Funding | Notable Differentiator | Source |
|---|---|---|---|---|
| WaffleVideo.Ai | Marketplace for custom, creator-sourced video data for AI training. | Pre-Seed / ~$250k (estimated) | Mobile-first creator workflow for 'Missions'; emphasizes rights-cleared, commercially safe footage. | [wafflevideo.ai] |
| Defined.ai | Marketplace for AI training data (text, image, audio, video). | Later stage / $100M+ total funding | Broad multi-modal data offering; established enterprise customer base. | [Crunchbase] |
| Wirestock | Marketplace for stock media (photos, videos, illustrations). | Seed / $1.1M (2023) | Traditional stock footage model; large existing creator community. | [Crunchbase, 2023] |
The competitive map splits into three distinct segments. First, generalist AI data marketplaces like Defined.ai offer a one-stop shop for text, image, audio, and video data, competing on breadth and enterprise-grade service [Crunchbase]. Second, specialized video data platforms, including Troveo AI and Protege, focus specifically on the video modality but often blend synthetic and real data sources [ProjectStartups, 2026]. Third, adjacent substitutes include traditional stock footage libraries (e.g., Wirestock, Shutterstock) and direct sourcing by AI labs, which provide volume but lack the purpose-built, 'mission'-specific attributes Waffle targets.
Waffle's current defensible edge rests on its creator acquisition and workflow design. The mobile app and structured 'Missions' system are built to elicit specific, commercially safe footage at scale from a distributed creator base, a process distinct from scraping the web or licensing existing stock libraries. This edge is perishable, however. It depends on executing a successful two-sided network launch before a well-funded competitor replicates the mission-based model or before synthetic video generation improves enough to reduce demand for certain types of real-world footage. The company's early association with the FoundersBoost accelerator and Hallstone Ventures provides initial validation but not a capital moat [ProjectStartups, 2026], [Startups.com, 2026].
The company is most exposed on the demand side. It has not publicly named any AI company customers, while competitors like Defined.ai list enterprise clients and Troveo AI has been covered in trade press [Crunchbase], [ProjectStartups, 2026]. Without demonstrated buyer traction, Waffle risks building a creator supply pool with no clear path to monetization. Furthermore, its 30% revenue share to creators, while a key incentive, must be balanced against platform fees and may pressure unit economics compared to synthetic data providers or firms with proprietary datasets [Perplexity Sonar Pro Brief].
The most plausible 18-month scenario hinges on Waffle securing a flagship enterprise partnership. If the company can sign and announce a deal with a recognizable AI lab or model developer within the next year, it would validate the demand for its custom footage and likely attract further creator supply and investor capital. In that case, specialized rivals like Protege or Kled AI, which also lack public traction, could lose momentum. Conversely, if no such partnership materializes and synthetic video quality continues its rapid advance, Waffle's value proposition for certain object and scene footage could diminish, making it a niche player while generalists like Defined.ai continue to consolidate market share.
Partially corroborated -- Competitor identification is sourced from a single aggregator; funding and differentiation for most named competitors are not independently verified.
Opportunity
Open sources The prize for WaffleVideo.Ai is a position as the primary supply-side clearinghouse for high-quality, rights-cleared video data, a foundational input for the next generation of multimodal AI models.
The headline opportunity is to become the default marketplace for custom video training data, a role analogous to what Getty Images represents for stock photography but purpose-built for AI's unique needs. The company's early positioning on creator economics and mobile-first workflow suggests a path to aggregating a large, diverse video library with clear commercial rights, a critical pain point for AI developers navigating copyright uncertainty. While the scale of this outcome is aspirational, the cited evidence of an active beta app and creator recruitment efforts indicates the initial supply-side motion is operational, a necessary first step toward a two-sided platform [Cornell career listing, Feb 2026].
Growth from this early stage could follow several concrete, non-mutually exclusive paths. The following scenarios outline plausible routes to significant scale.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Creator Network Dominance | Waffle becomes the go-to platform for millions of creators seeking micro-tasks, amassing a proprietary video library unmatched in size and diversity. | A viral creator marketing campaign or a partnership with a major creator platform (e.g., TikTok, YouTube) drives mass sign-ups. | The company's public messaging is heavily creator-centric, and its 30% revenue share model is designed to attract supply [Perplexity Sonar Pro Brief]. Active recruiting at university career centers suggests a targeted, low-cost acquisition strategy is already in motion [Cornell career listing, Feb 2026] [Quinnipiac University, 2026]. |
| Vertical Specialization | The company captures dominant market share in one or two high-value verticals (e.g., medical procedures, automotive manufacturing, retail environments) where data scarcity commands premium pricing. | Securing a flagship enterprise customer in a specific industry that commissions large, proprietary datasets. | The "Missions" system allows for directed data collection, a structure inherently suited to fulfilling specific, high-value client requests [Perplexity Sonar Pro Brief]. |
| Infrastructure Embedding | Waffle's data pipeline and licensing framework become an embedded component of major cloud AI platforms (AWS, Google Cloud, Azure) as a managed data service. | A strategic partnership or API integration with a cloud hyperscaler's AI/ML suite. | The focus on commercially safe, rights-cleared data directly addresses a major infrastructure gap for enterprise AI deployment, making it a potential value-add for cloud providers seeking to de-risk customer model development. |
Compounding for a marketplace of this type would manifest as a classic network effect, but with a data-quality twist. Each new creator expands the library's breadth and potential niche coverage. Each new AI company customer, in turn, generates more licensing revenue, a portion of which flows back to creators, reinforcing their loyalty and incentivizing further contribution. Critically, approved video clips that are licensed and then relicensed create a recurring revenue stream without additional creator effort, improving platform unit economics over time. The company's claim of a 30% creator share on relicensing is a direct, if unverified, acknowledgment of this intended flywheel [Perplexity Sonar Pro Brief].
The size of the win, should a dominant marketplace scenario play out, can be contextualized by looking at comparable data-centric platforms. For instance, Shutterstock, a publicly traded licensing marketplace for creative assets, held a market capitalization of approximately $1.4 billion as of early 2026. A platform that becomes essential for training the world's visual AI models,a potentially larger and more captive market than traditional stock media,could command a significant premium. If WaffleVideo.Ai captured a meaningful portion of the custom video data segment, a valuation in the hundreds of millions to low billions is a plausible outcome (scenario, not a forecast). This upside is what makes the early-stage risk profile worth examining in the private diligence section.
Partially corroborated -- Opportunity analysis based on company claims and operational signals; market comparables are public, but Waffle's specific path to scale remains unproven.
Sources
Open sources
[wafflevideo.ai] Waffle Video | Get Paid to Film Video for AI Training | https://wafflevideo.ai/
[Perplexity Sonar Pro Brief] WaffleVideo.Ai Brief | https://www.perplexity.ai/
[postPerspective, 2026] Waffle Video Profile | https://www.postperspective.com/
[ProjectStartups, 2026] Waffle Video Funding | https://www.projectstartups.com/
[Quinnipiac University, 2026] Paid Video Tasks for AI Training (Remote) | https://cas360.qu.edu/jobs/waffle-video-inc-paid-video-tasks-for-ai-training-remote/
[Grand View Research, 2025] Data Collection and Labeling Market Size Report | https://www.grandviewresearch.com/
[The Information, 2024] OpenAI's Sora Video Model | https://www.theinformation.com/
[Reuters, 2024] AI Copyright Lawsuits | https://www.reuters.com/
[Crunchbase] Defined.ai Profile | https://www.crunchbase.com/
[Crunchbase, 2023] Wirestock Funding | https://www.crunchbase.com/
[Startups.com, 2026] Waffle Video Accelerator | https://www.startups.com/
[Cornell career listing, Feb 2026] Paid Video Tasks for AI Training (Remote) | https://career.cornell.edu/jobs/paid-video-tasks-for-ai-training-remote-wafflevideo-ai
Articles about WaffleVideo.Ai
- WaffleVideo.Ai's Creator Marketplace Lands a Pre-Seed Check for AI Training Footage — The Los Angeles startup is paying creators a 30% revenue share for rights-cleared video clips, betting on a mobile-first workflow to supply AI companies.