Comet, which provides tools for data scientists and others to build ML models, raises a $50M Series B led by OpenView, bringing its total funding to almost $70M
As machine learning becomes a more integral part of running businesses, the model-building process still requires iteration and experimentation.
Context & Ripple Effects
Comet's $50M Series B closes a pivot arc: the company began in 2018 as a marketplace for tech and data freelancers with a $12.8M Daphni-led round, and has since repositioned around tooling that helps data scientists build and iterate on ML models. The new round, led by OpenView, brings its total funding to nearly $70M.
The timing matters: OctoML raised an $85M Series C from Tiger Global the same week, meaning two ML-lifecycle tooling vendors drew large rounds within days of each other — evidence that investors were funding the build-and-deploy layer around models, not just the models.
First-order effects
- Comet gains roughly $70M in cumulative backing and an OpenView partnership to scale its model-building and experimentation platform, while OpenView adds it to a portfolio bet on ML developer tooling.
- Data science teams evaluating experiment-tracking and iteration tools get a better-capitalized vendor whose pricing and roadmap can now compete at enterprise scale.
Second-order effects
- OctoML's near-simultaneous $85M raise forces the two companies to stake out opposite ends of the ML lifecycle — Comet on building and iterating, OctoML on optimizing and deploying — with each likely pushing toward the other's territory to own more of the workflow.
- Enterprise buyers gain leverage as funded vendors compete for the same ML-platform budgets, pressuring bundling of experimentation, optimization, and deployment into single contracts.
Third-order effects
- If the pattern holds through later rounds like Applied Compute's climb toward a $1.3B valuation, capital keeps consolidating in the complement layer around models — the tools that customize, iterate, and operationalize them — suggesting value accrues to whoever owns the workflow rather than the model itself.
- A well-funded tooling tier makes ML adoption cheaper for ordinary enterprises, widening the buyer base beyond AI-native firms and entrenching these platforms as standard infrastructure.
The trend: Venture capital is systematically funding the tooling and complement layer around machine learning models — building, iterating, deploying — as investors bet the workflow, not the model, is where durable value lands.