Open-source Omniconvert alternatives

PostHog is an open-source platform that bundles the tools product teams use to understand and improve what they ship. It captures event data, user sessions and errors in one place, and its newer self-driving mode aims to turn signals such as errors, rage clicks and failed queries into researched reports and pull requests that engineers review and merge.

The product suite includes product analytics with autocapture or manual instrumentation and SQL analysis, web analytics with a dashboard resembling Google Analytics, session replays for web and mobile, feature flags, experiments with statistical measurement, surveys, error tracking, logs, a data warehouse and a customer data platform. The back end is Python and the front end is React and TypeScript.

PostHog's repository license is listed as 'Other' on GitHub, so check which parts are open source and which are enterprise-only. It can be self-hosted, and the company also runs a managed cloud service. It suits product and engineering teams that prefer a single tool over separate analytics, flagging and replay vendors.

GrowthBook combines feature flagging, experimentation and product analytics in one open-source platform. The project's stated philosophy is that most companies cannot build an in-house platform of this kind, so it aims to give them the flexibility of one without the effort, instead of paying for closed SaaS tools or stitching together unmaintained libraries.

Feature flags support targeting, gradual rollouts and experiments, with 24 SDKs including React, Python, Android and iOS. The statistics engine offers methods such as CUPED, sequential and Bayesian testing, post-stratification, bandits and sample ratio mismatch checks. It is warehouse native, querying data sources such as BigQuery, Snowflake and Databricks, with metrics defined in flexible SQL. A built-in analytics suite supports dashboards, and there are webhooks, a REST API and an MCP server.

You can start with a free GrowthBook Cloud account or run it yourself with Docker and the prebuilt public image. The repository lists the licence as Other, so review the licence terms for the exact conditions on the open-source and enterprise parts. It suits product and data teams that want to run experiments on their own warehouse data.