Upscale AI
End-to-end AI platform for creating and running performance ads on Streaming TV.
Website: https://www.upscale.ai/
Cover Block
PUBLIC
| Field | Value |
|---|---|
| Name | Upscale AI |
| Tagline | End-to-end AI platform for creating and running performance ads on Streaming TV |
| Headquarters | San Francisco, United States |
| Founded | 2020 |
| Stage | Seed |
| Business Model | SaaS |
| Industry | Media / Entertainment (AdTech) |
| Technology Type | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Amit Mastud, Sandeep Singh, Yeshu Singh |
| Funding Label | Seed |
| Total Disclosed | ~$5.5M [BusinessWire, May 2025] |
Links
PUBLIC
- Website: https://www.upscale.ai/
- Product (Studio): https://studio.upscale.ai/
- Careers: https://www.upscale.ai/careers
- Whitepaper: https://www.upscale.ai/whitepaper
- PitchBook profile: https://pitchbook.com/profiles/company/464356-72
- Crunchbase profile: https://www.crunchbase.com/organization/upscale-579e
Executive Summary
PUBLIC
Upscale AI is a San Francisco company building what it describes as the first end-to-end AI platform for creating and running performance advertising campaigns on Streaming TV [Upscale AI, retrieved 2025]. The pitch combines two distinct workflows that have historically required separate vendors: generative AI for video creative production, and machine-learning-driven programmatic buying against connected TV (CTV) inventory [BusinessWire, May 2025]. The company was founded in 2020 by Amit Mastud, Sandeep Singh, and Yeshu Singh, with a team that the funding announcement describes as drawn from Google, Twitter, Salesforce, Moloco, and MNTN [BusinessWire, May 2025]. Upscale launched publicly in May 2025 alongside a $5.5 million seed round led by NVP Capital with participation from M12 (Microsoft's venture fund), Eniac Ventures, SuperAngel.Fund, and Breakpoint Capital [The SaaS News, May 2025]. The product, branded Upscale Studio, is positioned at commerce brands that want to treat TV more like a measurable performance channel rather than a brand-awareness line item [BusinessWire, May 2025]. The business model appears to be SaaS plus media spend management, with the careers page indicating an insertion-order-based onboarding motion [Upscale AI Careers, retrieved 2025]. Over the next 12 to 18 months the questions worth tracking are whether Upscale can demonstrate measurable ROAS uplift versus incumbents like MNTN and the major DSPs, whether brand customers will accept AI-generated video as primary creative, and how quickly the team converts beta users into paying logos.
Data Accuracy: GREEN -- Confirmed by BusinessWire, The SaaS News, PitchBook, and the company's own website.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | SaaS (with managed media component) |
| Industry / Vertical | AdTech / Connected TV |
| Technology Type | Generative AI + ML programmatic bidding |
| Geography | North America (HQ San Francisco) |
| Growth Profile | Venture Scale |
| Founding Team | Three co-founders, AdTech and big-tech veterans |
| Funding | $5.5M seed, May 2025 |
Company Overview
PUBLIC
Upscale AI was incorporated in 2020 according to PitchBook, but operated quietly until its public launch in May 2025 [PitchBook, retrieved 2025] [BusinessWire, May 2025]. The roughly five-year gap between formation and launch is consistent with a stealth build period, and the company unveiled its product (Upscale Studio) and seed financing in the same announcement, suggesting the team prioritized having a shippable platform before going to market [BusinessWire, May 2025].
The company is headquartered in San Francisco and is led by co-founders Amit Mastud, Sandeep Singh, and Yeshu Singh, whose backgrounds the launch release attributes to Google, Twitter, Salesforce, Moloco, and MNTN, all companies with relevant exposure to ad systems, programmatic infrastructure, or performance TV [BusinessWire, May 2025]. The Moloco and MNTN lineage is particularly notable for context: Moloco is a leading ML-driven programmatic platform for app and commerce advertisers, and MNTN built one of the first self-serve performance CTV products. Upscale's positioning sits at the intersection of those two playbooks.
Key milestones in the public record are limited but coherent: incorporation in 2020 [PitchBook, retrieved 2025], a private build period, the May 2025 launch of Upscale Studio, and the simultaneous $5.5 million seed round led by NVP Capital [BusinessWire, May 2025] [The SaaS News, May 2025]. The company is currently running a beta program based on the public sign-up page [Upscale AI, retrieved 2025].
Data Accuracy: GREEN -- Confirmed by BusinessWire, PitchBook, and the company website.
Product and Technology
MIXED
Upscale Studio is described by the company as an end-to-end workflow that takes a brand from creative generation through programmatic execution on Streaming TV inventory [Upscale AI, retrieved 2025] [PUBLIC]. On the creative side, the platform uses generative AI to produce video assets intended to meet the quality bar of broadcast-grade TV spots [BusinessWire, May 2025] [PUBLIC]. On the buying side, the company says it uses "state-of-the-art ML models to deliver performance outcomes over real-time programmatic bidding" across goals such as new customer acquisition, website visits, and audience engagement [Upscale AI, retrieved 2025] [PUBLIC]. The combination is what the company calls the first AI-native platform for performance TV [BusinessWire, May 2025] [PUBLIC].
The target customer profile, as inferred from the careers page and product copy, is commerce brands (likely DTC and mid-market retailers) that already run performance marketing on Meta and Google and want to extend the same measurement discipline to CTV [Upscale AI Careers, retrieved 2025] [PUBLIC]. The onboarding motion described in the open New Business Sales role involves signing an initial Insertion Order and helping the brand stand up creative assets on the platform, which suggests a hands-on white-glove model in the early innings rather than pure self-serve [Upscale AI Careers, retrieved 2025] [PUBLIC].
Technology-stack details are limited in public sources. The product is delivered through a web application at studio.upscale.ai and a separate app.upscale.ai login surface, suggesting at least two distinct application contexts (likely a creative studio and a campaign console) [Upscale AI, retrieved 2025] [PUBLIC]. The company publishes a security policy and privacy policy, both standard for a B2B SaaS platform handling brand assets and campaign data [Upscale AI, retrieved 2025] [PUBLIC]. Specific model providers, inventory partners, measurement integrations, and identity-graph partners are not disclosed in the public materials reviewed.
Data Accuracy: YELLOW -- Product positioning confirmed by BusinessWire and the company website; tech-stack and partner detail not publicly disclosed.
Market Research and Opportunity
PUBLIC
The market matters now because television advertising is in the middle of a structural shift from linear to streaming, and the buyer base is shifting with it from brand teams to performance teams.
Connected TV ad spend in the United States has been one of the fastest-growing segments of digital advertising for several years, driven by cord-cutting, the rise of ad-supported tiers on Netflix and Disney+, and the maturation of programmatic CTV exchanges. Because the structured facts contain no third-party TAM citation specific to Upscale, this report avoids quoting a sized number it cannot anchor; readers should treat US CTV ad spend (commonly reported by eMarketer and IAB in the high-tens of billions of dollars annually) as the relevant pool, and AI-assisted creative production as a fast-emerging adjacent layer.
The demand drivers Upscale is implicitly betting on are three. First, performance marketers entering CTV want the same closed-loop measurement and bid-time optimization they get on Meta and Google, which is the gap MNTN and Tatari built businesses to fill [BusinessWire, May 2025]. Second, the cost and cycle time of producing TV-quality video creative has historically been a barrier for mid-market brands; generative video models are now plausibly good enough to produce at least some performance creative variants without a full production shoot. Third, the proliferation of streaming inventory across dozens of apps and FAST channels makes programmatic buying with ML-driven optimization more valuable than manual planning.
Adjacent and substitute markets include the major DSPs (The Trade Desk, Amazon DSP, Google DV360), performance-CTV specialists (MNTN, Tatari, Strategus), and the AI video generation tool category (Runway, Synthesia, Google Veo, OpenAI Sora) which could in principle bundle a media-buying layer or be bundled into one. Regulatory pressure is moderate: privacy regimes (CCPA, state-level acts) constrain identity-based targeting, and AI content disclosure rules are emerging in some jurisdictions. Neither is an immediate barrier to entry, but both shape long-term defensibility.
| Sizing claim | Value | Source |
|---|---|---|
| Upscale seed round | $5.5M | [BusinessWire, May 2025] |
| US CTV ad spend (analogous market, widely reported) | reported in the tens of billions USD annually | industry trade press |
The takeaway: Upscale is targeting a real and growing pool, but the specific TAM for an integrated AI-creative-plus-performance-buying platform is not yet sized in any cited source, and investors should treat market-size claims for this exact wedge as directional rather than measured.
Data Accuracy: YELLOW -- Funding figure confirmed; market sizing relies on analogous categories rather than a named report tied to Upscale's specific wedge.
Competitive Landscape
MIXED
Upscale is positioned as an integrated alternative to a stack that brands currently assemble from a creative agency, a CTV specialist, and a measurement vendor.
No named competitors are present in the structured facts captured for this report, so this section is written as prose rather than a comparison table. The competitive map nonetheless has a clear shape based on the categories Upscale's launch materials reference [BusinessWire, May 2025] [PUBLIC].
The incumbent layer in performance CTV is anchored by MNTN, which one of Upscale's founders is reported to have come from [BusinessWire, May 2025] [MIXED]. MNTN built the canonical self-serve performance-TV product for direct-to-consumer brands, with attribution and audience tools designed to look familiar to a Meta or Google buyer. Tatari plays a similar role with a more managed-service flavor, and Strategus focuses on the mid-market agency channel. On the open programmatic side, The Trade Desk dominates premium CTV inventory access, and Amazon DSP is increasingly important given Prime Video's ad-tier launch. These platforms do not generate creative; they buy media against creative the customer brings.
A second layer is the AI video generation category: Runway, Synthesia, HeyGen, Google Veo, and OpenAI Sora. These tools produce video assets but do not buy media. A brand using them today still needs a separate DSP or CTV specialist to actually run the campaigns, and the creative produced still needs to be reformatted, trafficked, and measured. Upscale's bet is that fusing the two layers in a single workflow is a meaningful product wedge, especially for commerce brands that lack in-house production teams [PUBLIC].
Where Upscale has a defensible early edge, if the product delivers, is workflow integration: a brand can in principle generate a creative variant, traffic it, see performance, and iterate inside one console. That edge is perishable to the extent that MNTN, Tatari, or The Trade Desk can bolt on a generative video module via partnership or acquisition, and to the extent that the foundation-model providers (Google, OpenAI, Meta) make creative generation a commodity. Where Upscale is most exposed is inventory access and measurement credibility: incumbents have years of direct supply relationships and case studies, and brand performance teams are conservative buyers who require proof. The most plausible 18-month scenario has Upscale winning a wedge of mid-market commerce brands that find MNTN too expensive and the DIY stack too fragmented ("winner if Upscale lands 50-plus paying brands with documented ROAS"), and losing ground if a larger DSP ships a competing AI-creative module bundled into existing seats ("loser if The Trade Desk or Amazon ship comparable generative tools natively").
Data Accuracy: ORANGE -- Competitor identification is inferred from the founding team's prior employers and category context rather than from sources naming Upscale's direct competitors.
Opportunity
PUBLIC
If Upscale executes, the prize is becoming the default performance layer for the next generation of TV advertising.
The headline opportunity. The single largest outcome for Upscale would be becoming the category-defining AI-native platform for performance Streaming TV, occupying for CTV the position MNTN built in the prior cycle but with a structurally lower cost base because creative production is automated rather than outsourced [BusinessWire, May 2025]. The reason this is reachable rather than aspirational is that the customer behavior already exists: commerce brands have been trained by Meta and Google to expect closed-loop performance measurement, and they have been trained by the recent generative-AI cycle to expect cheaper, faster creative iteration. Upscale's wedge is the integration of those two expectations into one workflow on a channel (CTV) where neither is yet standard [Upscale AI, retrieved 2025].
Growth scenarios.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Mid-market commerce land-and-expand | Upscale wins 100-plus DTC and mid-market retail brands on a self-serve plus light-touch model | Beta-to-paid conversion through 2025-2026 [Upscale AI, retrieved 2025] | The customer profile already buys performance media daily and is actively looking for CTV solutions [BusinessWire, May 2025] |
| Agency channel partnership | A holding-company agency or independent performance shop adopts Upscale as its default CTV creative-and-buying layer for a roster of brands | A signed agency partnership announcement | The team's MNTN and Moloco lineage gives it credibility with agency CTV leads [BusinessWire, May 2025] |
| Strategic acquisition target | A larger DSP, retail-media network, or streaming platform acquires Upscale to bolt on AI creative | M12's involvement gives Microsoft a natural look [The SaaS News, May 2025] | Adtech consolidation has historically rewarded integrated creative-plus-buying tools |
What compounding looks like. The flywheel for an integrated AI creative and buying platform is data: every campaign run on Upscale produces paired (creative variant, audience, outcome) data that can train the next generation of both the creative model and the bidding model. Over time, that data asset is the moat, because a generic foundation model does not know which six-second hook converts a household-goods buyer on Hulu at 9pm. Distribution compounds in parallel: once a brand has trafficked assets, audience definitions, and measurement pixels inside Upscale, switching costs rise. The careers page's emphasis on insertion-order-based onboarding suggests the company is already building the operational scaffolding for this kind of customer lock-in [Upscale AI Careers, retrieved 2025].
The size of the win. A credible public comparable is MNTN, which built a multi-billion-dollar private valuation in the same broad category before its 2025 IPO filing, widely covered in the trade press. Translating that into Upscale terms (scenario, not a forecast): if Upscale captures a meaningful share of the mid-market performance CTV wedge over the next five to seven years and AI-generated creative becomes accepted as a primary format, an outcome in the high-hundreds of millions to low-billions of dollars of enterprise value is in the range of plausible comparables for the category. That outcome requires Upscale to convert beta interest into paying brands, demonstrate ROAS, and survive the period in which incumbents will almost certainly ship competing AI creative features.
Data Accuracy: YELLOW -- Scenario framing is grounded in cited launch materials and category comparables; specific outcome sizes are explicitly labeled as scenarios rather than forecasts.
Sources
PUBLIC
[BusinessWire, May 2025] Upscale AI Launches With $5.5M in Funding and Unveils 'Upscale Studio' | https://www.businesswire.com/news/home/20250510370327/en/Upscale-AI-Launches-With-$5.5M-in-Funding-and-Unveils-Upscale-Studio-the-First-AI-Native-Platform-to-Create-and-Run-Performance-TV-Ads-on-Streaming-Channels
[The SaaS News, May 2025] Upscale AI Raises $5.5 Million in Funding | https://www.thesaasnews.com/news/upscale-ai-raises-5-5-million-in-funding
[PitchBook, retrieved 2025] Upscale.AI 2025 Company Profile: Valuation, Funding & Investors | https://pitchbook.com/profiles/company/464356-72
[Upscale AI, retrieved 2025] Upscale AI homepage | https://www.upscale.ai/
[Upscale AI, retrieved 2025] How it works | https://www.upscale.ai/how-it-works
[Upscale AI Careers, retrieved 2025] Careers | Upscale AI | https://www.upscale.ai/careers
[Upscale AI, retrieved 2025] Join Beta | https://www.upscale.ai/join-beta
[ZoomInfo, retrieved 2025] Upscale AI - Overview | https://www.zoominfo.com/c/upscale-ai-inc/544620653
[Crunchbase, retrieved 2025] Upscale - Crunchbase Company Profile | https://www.crunchbase.com/organization/upscale-579e
Articles about Upscale AI
- Upscale AI Wants to Turn Every Streaming TV Ad Slot Into a Performance Buy — The San Francisco startup raised $5.5M from NVP Capital and Microsoft's M12 to make CTV behave like Google Ads.