Process Innovation Challenge, Round 1

Home/Process Innovation Challenge, Round 1

Process Innovation Challenge, Round 1
Track: Advanced Localization Management | AL4 |
Thursday, October 18, 2018, 4:30pm – 5:15pm
Held in: Harbor
Presenters:
Konstantine Boukhvalov - Experis-ManpowerGroup 
Rikkert Engels - Xillio 
Dalibor Frívaldský - Memsource 
Patrícia Paladini Adell - CA Technologies 
Konstantin Savenkov - Intento, Inc. 
Vincent Swan - Pactera Technologies
Moderator: Jeff Kiser
Hosts: Kerstin Bier,
Yuka Ghesquière Nakasone,
Jonas Ryberg

Shortlisted process innovators will pitch their innovative process ideas to a panel and the audience. Any method of getting their idea across is supported, but the clock is ticking and the Process Dragons and audience will have their questions. Fast and furious is the pace, and only attendees and the panel will vote for their favorite. The top innovations will go through to the final round on day two, to be named as the LocWorld Process Innovator for North America 2018.

Localization Needs an Updated Connector Strategy – Rikkert Engels (Xillio)
Xillio developed a content integration platform as a service solution providing 20+ connectors that are all exposed through one single API. These connectors provide features that allow for continuous localization and in context.

FlexData – Enriching TermBase Architecture – Konstantine Boukhvalov (Experis-ManpowerGroup)
The proposed solution is based on the approach where a single TermBase entry has multiple active values available for matching in a computer-assisted translation environment. Combining FlexData multidimensional TermBase entry approach with existing fuzzy matching will result in optimization of both TermBase management and translation process.

Using AI in Human Workflows – Konstantin Savenkov (Intento, Inc.)
Human-to-artificial intelligence (AI) tools are used so that language service providers may play with machine translation (MT) and optical character recognition at scale before going ahead with the API integration. Command-line and web interfaces try to evaluate multiple MT engines at once, both stock and custom.

Complementing TM with AI-based NTs and MT QE – Dalibor Frívaldský (Memsource)
Artificial intelligence (AI)-powered non-translatables (NTs) more accurately recognize segments that don’t require translation and provide a confidence score from 99%-100%. Machine translation (MT) quality estimation (QE) builds on this by identifying MT outputs that require little to no post-editing (85%-100%). Together with translation memory (TM), these innovations bring translation efficiency to new levels.

Automating and Simplifying the Translation Allocation Workflow Using Repurposed Marketing Tools – Vincent Swan (Pactera Technologies)
As part of our optimization process during deployment of our OneForma production ecosystem, Pactera Labs has developed a simplified process to facilitate job notification, allocation and tracking for certain program profiles. This is facilitated without requiring complex translation management systems or other workflow management tools. It supports our resource management and analytics processes and significantly reduces project management and freelancer overhead by simplifying the job acceptance routine and providing automated dynamic status tracking.

Shift Left on Internationalization Readiness – Patrícia Paladini Adell (CA Technologies)
A proposed system that automates the following tasks in the localization process: resource files validation for internationalization readiness, pseudolocalization, user interface screen for in-context translation and the creation, leverage and route of translation packages generated from development resource files.