Luminance is a legacy platform rooted in e-discovery. It is a heavy-duty tool designed to visualise statistical patterns across massive datasets. Colabra is a modern M&A workspace. While Luminance focuses on complex 3D visualisations and unsupervised clustering, Colabra focuses on the deal workflow: turning files into a risk register, gap analysis, and entity map without the steep learning curve.
Luminance is known for its "star field" visualisation—a 3D cloud of documents grouped by similarity. While visually striking, users often find it difficult to navigate in practice. It requires significant manual tagging to "teach" the system what matters. It shows you that a document is an outlier, but leaves it to you to figure out why.
Colabra is designed for the deal team, not the data scientist. We do not show you a star field; we show you a risk register. We strip away the visual noise to focus on the signal: which documents are missing, which clauses are off-market, and which entities are high-risk.
Luminance detects deviation. Its "traffic light" system tells you that a contract is 10% different from the others. This creates false positives—a contract might be "different" because of formatting, not risk. Colabra detects legal concepts. We flag a document because it contains a specific liability risk, not just because it looks different from its neighbours.
Luminance is often described as a "project" to set up. It relies on the user to tag documents and train the model on each deal. Colabra comes pre-loaded with the buy-side playbook. You do not need to train the model; it already knows what a "change of control" clause looks like.
Luminance relies on unsupervised machine learning (clustering). It is opaque—often leaving lawyers wondering why the AI grouped certain files together. Colabra is transparent. Every flag we raise is linked to a specific clause, with a direct citation you can verify in seconds.