API Technical Specs
Learn how to configure and leverage our services to achieve your toughest recruiting needs.
Technical Specifications
AI Matching Engine
Sovren has the most robust AI matching engine because we’ve spent decades perfecting it. Our engine’s technical specifications are detailed below so you can see exactly what to expect when using our premier matching technology — from integration to output.
Deployment | 100% Cloud-hosted solution. |
Integration | Available as a cloud-based service via a REST API. Official, open-source SDKs are available for some languages. |
Stability | Built-in monitoring to automatically detect and repair corrupted indexes. Data is automatically partitioned for performance and archiving. |
Scalability | Runs on Amazon AWS, scales both vertically and horizontally and has no document limits. |
Input Formats | Parsed document input supported in JSON or XML format. |
Compatibility | The Sovren AI Matching Engine works with the Sovren Resume Parser and the Sovren Job Order Parser to provide advanced matching capabilities between a source document and candidates. |
Distinctive Features | Our engine leverages our proprietary AI Matching algorithm that uses profiles rather than keywords. It's the only engine that allows hybrid Boolean/Semantic searches. |
Configurable Categories | Each matching transaction has a user-configurable set of categories that influence the overall score, which allows the engine to be fine-tuned to better serve applications across multiple industries. Categories include:
|
Transaction Output | Output includes a list of results ordered in descending order of the overall score. Scores are broken out by category for each match and include analysis messages that describe a transaction. |
Matching Directions | Resume : Job (Find the best matching jobs for a resume) Job : Resume (Find the best matching resumes for a job) Resume : Resume (Find resumes similar to this one) Job : Job (Find jobs similar to this one) |
Semantic Searches | Can be built using an object, or queries can be hand-rolled (actual query strings). |
Data Replication | Replicates your most recent data for performance and high availability. |
Speed | Sub second matching/searching transaction times. |
Scoring | Our AI Matching Engine provides absolute scoring, where the best match is reported at its true score from 0-100. We also report sub-scores by different categories of data. |
Search Filtering | Supports user specified data fields for searching/filtering. |
Semantic Indexing of Data | Data Indexed:
|