Welcome!

The e2elink executes a full record linkage pipeline without the need for human intervention.

Matching of clinical data made easy

Matching clinical entries from two datasets is an arduous task that requires extensive data curation. We have simplified this, finally.

Automated steps

The manual process of record linkage has a set of well-defined steps. We have automated all of them.

1. Schema matching

2. Data preprocessing

3. Blocking

4. Comparisons

5. Scoring

6. Evaluation

Powered with machine learning

We have generated millions and millions of synthetic data points. This means that we have run this pipeline against a large number of scenarios, and we can learn from it. The end-to-end record linkage pipeline works because we have trained machine learning models based that can guide us through the real case scenarios.