CausalML
CausalML is an open-source Python package developed and maintained by Uber that provides a suite of machine learning methods for causal inference and uplift modeling.
Profile
CausalML provides Python tools to estimate the causal effect of treatments or interventions from data, helping businesses optimize decisions like which customers to target with an offer.
CausalML is an open-source Python package developed and maintained by Uber that provides a suite of machine learning methods for causal inference and uplift modeling. Originally released in 2019, the project has accumulated over 5,900 stars on GitHub and 855 forks, with 680 commits as of June 2026. The library is designed to help data scientists and engineers estimate the causal effect of treatments or interventions from observational and experimental data, using techniques such as meta-learners, tree-based methods, and neural networks.
It is not a standalone company; it remains a project under Uber's open-source umbrella, with no disclosed revenue, funding, or formal customer base. The most recent commit, made on June 5, 2026, addresses compatibility with scikit-learn v1.9.0, indicating ongoing maintenance. The project's documentation and release notes show it has evolved through 16 versions, with the latest v0.16.0 upgrading to manylinux_2_28 and removing scipy version constraints.
CausalML competes with other causal inference libraries like DoWhy, EconML, and CausalNex, but its primary differentiator is its focus on uplift modeling for business applications such as marketing, pricing, and policy evaluation. The project has no known controversies, leadership turnover, or customer churn, as it is not a commercial entity. Its primary risk is dependency on Uber's continued investment; if Uber deprioritizes the project, maintenance could slow. The library is used by data science teams at various organizations, but no specific named clients are disclosed in the available sources.
Who buys this
- Data science teams at large technology companies needing uplift modeling for marketing campaigns
- E-commerce and retail firms optimizing pricing and promotions
- Healthcare organizations evaluating treatment effectiveness from observational data
- Financial services companies assessing the impact of policy changes on customer behavior
- Academic researchers studying causal inference methods
Strengths and what to watch
Strengths
- Comprehensive library of uplift modeling algorithms, including meta-learners, tree-based methods, and neural networks, all in a single package
- Strong community adoption with over 5,900 GitHub stars and 680 commits, indicating active use and contribution
- Backed by Uber's engineering resources, ensuring ongoing maintenance and compatibility with modern Python and scikit-learn versions
Watch for
- As an open-source project under Uber, its long-term viability depends on Uber's continued investment; if deprioritized, maintenance could slow
- No disclosed revenue, funding, or formal customer base, making it difficult to assess commercial traction or market share
- Competition from other causal inference libraries (DoWhy, EconML, CausalNex) may fragment the user base and reduce adoption
Key Information
- Industry
- MLOps & AI Infra
- Founded
- 2019
Frequently Asked Questions
What is CausalML?
CausalML is an open-source Python package developed by Uber that provides machine learning methods for causal inference and uplift modeling. It helps data scientists estimate the causal effect of treatments or interventions from observational and experimental data.
What is uplift modeling in CausalML?
Uplift modeling in CausalML focuses on estimating the incremental impact of a treatment, such as a marketing offer, on an individual. It uses meta-learners, tree-based methods, and neural networks to identify which customers are most likely to respond positively.
How does CausalML compare to DoWhy or EconML?
CausalML differentiates itself by specializing in uplift modeling for business applications like marketing and pricing. While DoWhy and EconML focus on broader causal inference, CausalML provides a comprehensive suite of uplift algorithms in a single package.
Is CausalML still maintained by Uber?
Yes, CausalML remains under Uber's open-source umbrella. The most recent commit on June 5, 2026, addressed compatibility with scikit-learn v1.9.0, and the project has evolved through 16 versions, indicating ongoing maintenance and updates.
What are the main strengths of CausalML?
CausalML offers a comprehensive library of uplift modeling algorithms, strong community adoption with over 5,900 GitHub stars, and backing from Uber's engineering resources. This ensures ongoing maintenance and compatibility with modern Python and scikit-learn versions.
What are the risks of using CausalML?
The primary risk is that CausalML's long-term viability depends on Uber's continued investment. If Uber deprioritizes the project, maintenance could slow. Additionally, competition from other causal inference libraries may fragment the user base.
Sources
- github.com — Project details: 5.9k stars, 855 forks, 680 commits, latest commit on 2026-06-05 for sklearn v1.9.0 compatibility, v0.16.0 release notes
- www.asml.com — ASML Q1 2026 financial results (€8.8B net sales, 53% gross margin) — used to confirm the date context and that the dossier contains unrelated company data
- investor.figma.com — Figma Q1 2026 financial results ($333.4M revenue, 46% YoY growth) — used to confirm the date context and that the dossier contains unrelated company data
- investor.marvell.com — Marvell Q4 FY2026 financial results ($2.219B revenue, 22% YoY growth) — used to confirm the date context and that the dossier contains unrelated company data
- grail.com — GRAIL Q1 2026 financial results ($40.8M revenue, 28% YoY growth) — used to confirm the date context and that the dossier contains unrelated company data
- techcrunch.com — TechCrunch article on Q1 2025 startup funding records — used to confirm the date context and that the dossier contains unrelated company data