Fundraising Fox

Womp Labs

Entrepreneur First '24

New York, US · Founded 2024 · 1 known investors

womplabs.ai ↗

Womp Labs builds diagnostic tooling that pinpoints where machine learning models underperform, surfacing the specific data cohorts and features responsible. Machine learning teams use it to correct training data and address reliability problems before a model is put into production.

Also known as womplabs

Developer ToolsEnterprise Software

Founders & leadership

Womp Labs was founded in 2024 by Joe Melkonian and Eric Sun.

JMJoe Melkonian
Joe MelkonianinCo-Founder and CEO2024–2025Joe Melkonian reached the finals of ISEF with research on efficient data attribution. He went on to become a founding engineer at Insight Cyber, where he developed machine learning systems used for anomaly detection across Fortune 500 and government networks.
ES
Eric SuninCo-Founder2024–2024Eric Sun is the engineer behind what is recognized as the world's most memory-efficient real-time operating system, and he has also worked on low-latency system design at Infineon and high-performance computing pipelines at Citadel. Earlier, he earned three gold medals at Science Olympiad Nationals.

Investors · 1

Company profile

researched Aug 2026

Womp Labs is an early-stage company working in data infrastructure and analytics for machine learning. Public profiles describe its products as AI diagnostic tools that identify the data cohorts and feature-level issues responsible for degraded model performance, with the goal of helping teams refine training data and improve model reliability prior to deployment.

One competitive-intelligence profile characterizes the company's positioning as making data "as causal and adaptable as code," emphasizing data quality as the lever for scaling AI performance and surfacing links between training data and observed model behaviour. Directory listings tag the company across AI infrastructure, MLOps, data platform, foundation model and machine learning categories, and describe it as a B2B business.

The company maintains an organization account on Hugging Face under the handle "womplabs" with three listed team members and ten public datasets, most recently updated in October 2024; no public models are listed there. Sources indicate a headcount in the 1-10 range.

Business model

Directory data classifies Womp Labs as a B2B business selling into machine learning and AI engineering teams; no pricing, contract structure or revenue details are disclosed in the available sources.

Traction

Publicly verifiable traction signals are limited: a headcount of 1-10, three Hugging Face team members, and ten public datasets published in late 2024. No customers, revenue or funding figures are disclosed in the sources reviewed.

Latest developments

The most recent dated public activity in the sources is Hugging Face dataset updates in October 2024. One directory (StartupHub.ai) lists the company's current operating status as "Inactive," while other listings continue to present it as an operating portfolio company; the sources do not explain or corroborate the inactive designation.

▸Full profile — market position, technology, go-to-market, geography, history, risks & controversies

Market position

Listed among new entrants in the MLOps sub-sector on the StartupHub directory, which assigns the profile an overall score of 22 (top 23% of profiles), with component scores for team (60), profile (50), agent readiness (21), visibility (10), community (8) and traction (0). The company is also part of the Entrepreneurs First portfolio.

Positioning centres on treating data as causal and adaptable, uncovering explicit links between specific training data and model behaviours rather than reporting aggregate model metrics.

Technology

Diagnostic tooling for machine learning models that isolates data cohorts and features associated with performance degradation and surfaces links between training data and model behaviour. The company has published ten public datasets on Hugging Face, ranging in size from roughly 1.3 thousand to 423 thousand rows, updated in September and October 2024.

Go-to-market

Machine learning and AI teams responsible for training data quality and model reliability before deployment.

Geography

Headquartered in New York, United States.

History

Sources place the company's founding in 2024 in New York, United States. A Hugging Face organization account was active with dataset uploads between September and October 2024. Womp Labs appears in the Entrepreneurs First portfolio listing.

Risks & controversies

Conflicting status information across sources: StartupHub.ai marks the company's operating status as "Inactive" without supporting detail, while the Entrepreneurs First portfolio and a competitive-intelligence profile present it as an operating company. Public information overall is sparse, with no disclosed funding, customers or product documentation in the available sources.

Compiled by commissioned research from 4 cited public sources — announcements, filings, and press listed under research sources below.

Key figures

latest reported
Employee count rangeJan 20251-10 employees
Estimated tech spendJan 2025$160
Hugging face team membersOct 20243 people
Public datasets on hugging faceOct 202410 datasets
Startuphub total scoreJan 202522 score (0-100)

Company-reported or press-reported figures, each dated to when it was claimed — not independently audited.

▸Research sources · 4

primary sources listed

4 public sources were cited for this profile; the first-party ones are listed here.

Frequently asked questions

What does Womp Labs do?
Womp Labs is a New York-based AI diagnostics startup that pinpoints data cohorts and features degrading model performance.
Who founded Womp Labs?
Womp Labs was founded by Joe Melkonian, Eric Sun in 2024.
Who are Womp Labs's investors?
Womp Labs's investors include Entrepreneur First.
Where is Womp Labs headquartered?
Womp Labs is headquartered in New York, US.