Public company, workplace, funding, and market signals
Updated Jul 29, 2026
YC-backed San Francisco startup building AI evaluation and human-data infrastructure that auto-generates bespoke annotation pipelines, routing, and QA for custom evals and post-training data.
Primary product
Long-horizon RL environments / AI evaluation and annotation platform
Founded
2025
Headquarters
San Francisco, California, United States
Team size
1-10
Industry
Software Development
Sub-industry
AI evaluation and annotation platform
Offices
Business model
Stage
Seed
Total raised
$500K
Latest round
Seed · Oct 2025
Latest amount
$500K
Oct 2025 · Undisclosed (top-tier VCs)
Investors
Founder-led, technical, and mission-driven; emphasizes measurable model improvement, rapid iteration, and hiring software engineers, researchers, and operators.
Compensation
Public hiring materials show a Founding Engineer role at $100K-$200K base salary plus 0.50%-1.50% equity; no broader company-wide compensation policy was publicly disclosed.
Pricing
Subscription; public third-party pricing pages cite a $29/mo starting price.
Differentiators
Technology
Customers
Competitors
Estimated revenue
Approximately $0.3M-$0.6M ARR (third-party estimates)
Estimated monthly visits
702
aVenture · Jun 2026
A recent company profile summarizes MangoDesk as a San Francisco startup founded in 2025 with about 2 employees, a pre-seed/seed financing profile, and a product focused on bespoke data-annotation pipelines and AI evaluation workflows.
LinkedIn · Aug 2025
The launch post presents MangoDesk as infrastructure for evals and human data pipelines, says it generates complete annotation pipelines from natural-language specs, and notes the founders’ Scale AI and Fortune 500 backgrounds.
Y Combinator
YC’s company page and launch copy describe MangoDesk as a Summer 2025 company building long-horizon RL environments and a natural-language system for generating custom annotation pipelines, with Ankith Subramanya as CEO and Ananth Subramanya as CTO.
Extruct AI
Third-party analysis describes MangoDesk as building production-grade RL environments for real workflows and estimates funding at $500K with a small San Francisco team.