At Affinda, our data science team are constantly looking at ways to improve the performance of our AI models to increase the accuracy that we can deliver to our customers. We use the latest developments in AI technologies to seek step-changes in performance, while we continue to fine-tune our models over time to drive improvements.
To ensure that we maintain our standard of high performance, we only make updates to the models that customers use when we are satisfied that the new model increases the overall accuracy of the parsed results. Customers should generally expect an increase in accuracy over time, however, this does mean that results can sometimes change for a given document, and our platform shouldn't be expected to produce exactly the same result every time the same resume is parsed across time.