This September we have details on AI-driven product updates designed to improve MT output, our latest partnership with Smartcat for stronger language services, and more community insights to let you in on. Every month at Intento we are having discussions around the next game-changing breakthrough, and this is your chance to get in the loop on the latest industry innovations. Whether you are a user, provider, or academic, there is something to be found in September 2021 Digest — so read on!
🚀 Groundbreaking MT quality of ICU strings — in any context
💪 Intento and Smartcat partner to strengthen the translation landscape
🤖 Introducing Intento Translator for Jira — take a look, global dev teams
🗣️ Let Chinese history speak to you with advanced MT for Classical Chinese
✍️ Use DeepL glossaries through Intento
🗓️ Keeping up with new opportunities on the event calendar
📚 Recommended reading for community insights
Groundbreaking MT quality of ICU strings — in any context
The AI team at Intento has been busy making strides to improve the quality of your MT output with context-aware ICU translations. No matter what provider you are working with, Intento makes sure that your ICU strings are properly translated. Considering that professional translators will find themselves confused by the ICU format, this is a massive step towards raising the level of your MT past even the ‘gold standard of human translation. Let us know if you have trouble using MT with other sophisticated file formats — we’ll have a solution.
Intento and Smartcat partner to strengthen the translation landscape
This collaboration will benefit the entire translation community, drawing from the combined integrations and expertise of both industry leaders. The Intento MT Hub for Smartcat gives users access to the best-possible translations through the power of over 40 leading machine translation providers such as Google Cloud AutoML, Amazon Translate, and ModernMT. At the same time, those taking advantage of Smartcat’s seamless, single-window localization processes will further benefit from full transparency on machine translation usage and analytics.
Intento Translator for Jira ensures flawless multilingual communications in Jira Data Center
The Intento Translator for Jira is setting the new standard of collaboration between global dev teams by translating tickets, comments, and reviews in real-time, directly in Jira. With Intento’s AI-enabled technology, all you need to decide is which content and how you want to translate. New or existing tickets, whole tickets, or specific frames, automatically or manually, any way you need them, just without the hassle of translation quality or re-formatting. It’s simply the safest and most effective way to bring your global dev team closer together while saving time and increasing productivity.
Let history speak to you with advanced MT for Classical Chinese
Classical Chinese, also known as Literary Chinese (Izh), is now supported by both Microsoft and Baidu, giving both modern China and the rest of the world unique access to traditional Chinese culture. With the right machine translation and fine-tuning, tourists, students, and researchers can now have seamless access to this wealth of resources. Use Intento to facilitate translations from literary Chinese to modern forms of Chinese and other 190 languages offered on our platform.
Use DeepL glossaries through Intento to curate better productivity for your team
Spending too much time cleaning up MT results? All users can now apply their DeepL glossaries when working through Intento to increase custom options. Glossaries decrease the amount of time and effort spend in post-editing by handling your organization’s specific terminology. DeepL currently supports glossaries for English to/from German, French, and Spanish, and you can apply them to Intento after creation in the DeepL API.
Intento among language industry leaders at SlatorCon Remote
Wonder what all the buzz is around content generators like GPT-3? Earlier this month, Intento CEO Konstantin Savenkov sat down with fellow industry leaders Michel Lopez of e2f and Jochen Hummel of Coreon and ESTeam to discuss the content generation chain and new frontiers for development. If you couldn’t make the event you can catch up on what you might have missed here, and you can get in touch to learn more about our thoughts on the future of content generation!
Learn how to make terminology a provider agnostic feature at GALA Connected
Make sure to reserve your spot at GALA Connected 2021 so you don’t miss Intento’s Pavel Doronin presentation, ‘Do MT models dream of your corporate terminology?’. Making terminology a provider-agnostic feature is a challenge. You can compile glossaries from central storage for each specific provider, use tags for applying terminology or set up an automated post-editing where neural-powered terminology correction of the MT output is involved. We did all these things, and we want to present the benefits and drawbacks of them.
Tune into the Virtual TAUS Data Summit for info on large-scale, multi-domain MT evaluation
We’ll be presenting on ‘large-scale, multi-domain MT evaluation’. This is a great opportunity to get behind the driver’s seat of machine translation by learning everything you need to know about choosing the best-fit engines for your language pairs, industry sector, and content type. Stay tuned for more details to come!
Join us for a conversation around building the right MT program — the first time
MT program development is a high priority for most if not all localization managers. But making sense of the MT landscape is hard. Every vendor seems to have an MT offering — so where do you start and who can you trust? Intento CEO Konstantin Savenkov will be leading a panel of innovation and localization leaders from Agilent, Amadeus, AstraZeneca, and IKEA in a discussion around the biggest challenges they faced during their MT program development, lessons learned, best practices, successes, failures, and future goals.
Globalsaké presents roundtables on 6 unique international challenges
Key updates and insights from the AI/MT landscape
Follow these links for more industry updates:
1. Youtube now lets you translate comments instantly (at your own risk)
2. Announcing AI21 Studio and Jurassic-1 language models
3. 77-atom molecule key to brain-like computing architecture
4. Fine-tuned language models are zero-shot learners
5. Program synthesis with large language models
6. The onset of the super-cycle at SlatorCon Remote
7. Zoom opts for speech-to-text in new transcription and translation features
