Hopper Release
Rishi Khullar | May 7, 2019
Did you know we name all of our releases after famous heretics in history? This month, we’ve chosen Grace Brewster Murray Hopper, an American computer scientist and United States Navy rear admiral. She continued to work in computing after the war, leading the team that created the first computer language compiler, which led to the popular COBOL language.
Auto-Populate Fields Via Extraction Analysis
Our customers love the send-to-field functionality in the Heretik Viewer. This allows reviewers to select text and click a button to populate a field while creating a location-aware hyperlink between the field and the text. This is especially useful for quality control (QC) workflows where a second pass reviewer can not only see how a field was coded, but also why it was coded with the exact text that triggered the decision.
While this is manual, we are seeing savvy customers make great use of Regular Expression creation in the Heretik Viewer to pinpoint critical data points. These customers are finding that good regular expressions will find one or two results that will almost always contain the value that a field requires.
With this new feature, you can create a field, run Extraction Analysis via your regular expression, and populate the first-found result to that field across an entire saved search. Importantly, this will also give each field the send-to-field treatment with a location-aware hyperlink between the field and text.
Re-Usable Heretik Analysis Templates
We are introducing the concept of Analysis Profiles to bundle and re-use models for doing different types of Heretik Analysis.
You could, for example, create an Analysis Profile that uses Heretik Classification, Heretik Segmentation, but also runs Extraction Analysis via numerous custom regular expressions. When creating a new Analysis Set, you could select this profile and Heretik will run everything as part of the same job.
In the future, when we provide the ability for you to create your own custom models, you will be able to slice and dice these re-usable templates in myriad ways.
Faster Analysis and Earlier Insights
We are committed to making Heretik Analysis a faster and smoother experience with early and deeper insight into the status of jobs.
In this release, we have:
- Made Heretik Analysis roughly ~50% faster
- Removed all cases of “stuck” analysis jobs by adding a timeout for troublesome documents that are taking too long to analyze
- Removed memory issues that may cause errors during Heretik Analysis
- Added helpful information during analysis like the amount of time it’s taking, model-level status, and the number of items found such as sections, for example
- Added three new fields at the Master Agreement level for Number of Sections, Section Headings Found and Section Types Found. You can now report on sections at the Master Agreement level and you can also keep these fields up-to-date via Relativity Scripts that the Heretik application ships with
Manual Section Edit Workflow
We’ve overhauled Section Edit Mode to make the QC workflow for Segmentation much faster and more intuitive. We’ve removed several un-needed clicks and we’ve also made the experience far more obvious so that training new team members on this functionality should be much simpler.
SHARE THIS ARTICLE:
RISHI KHULLAR
Rishi is the Director of Product at Heretik. Prior to Heretik, he worked as a product manager at Relativity. Rishi began his career as a lawyer and worked as a judicial clerk for the Virgin Islands.
