MT Technology: Language Accessibility by Juan Duvanced https://padlet.com/juan39/hbwu2aiq655c Use the latest MT technology to address the global language students, professionals and organizations. en-us 2018-09-30 21:53:36 UTC 2018-10-30 12:15:43 UTC hello@padlet.com https://padlet-assets.s3.amazonaws.com/icons/Folder.png Current Technologies juan39 https://padlet.com/juan39/hbwu2aiq655c/wish/287390848 1- Alibaba Cloud (MT)
For e-Commerce and very general texts. It works really bad with most Roman languages as it uses BLEU scores. It needs a full MTPE and QA process after populating.
2- DeepL
It's a revolutionary API based on an independent MT patent that has a high-quality output for politics and society, finance, energy compliance policies, and even some technical topics.
3- Microsoft
4- SAP
5- Systran
It is focused on full website and simple and rich text translation.
6- Youdao
7- Amazon
8- Google Cloud (Google Translate; Neural)
9- SDL Language Cloud
10- Baidu
11- Yandex

Top reference:

]]> 2018-09-30 21:53:36 UTC https://padlet.com/juan39/hbwu2aiq655c/wish/287390848 Possible Products or Services juan39 https://padlet.com/juan39/hbwu2aiq655c/wish/287390849 1- Education platform for businesses with a paid subscription format (a demo would be available).
  • Lessons that will teach CTOs, CEOs, Localization Managers and everyone involved in the process, to properly understand the MT universe and make better and informed decisions.
  •  These will have 24/7 support in many languages and private 1-on-1s in case they need more details and examples according to their own situation.

Possible names: DIFEMTI! (Don't Fear Machine Translation!)
KMT Academy


2- In-house or over-the-air training for specific projects or for the general implementation of the right MT technology.

The principle is very similar to the educational platform but in this case, there is a deep customized approach. The name and logo could even be the same to my current business.

3- Deliver translations services fully powered by MT algorithms with a comprehensive MTPE and QA process. 
  •  The user will go through a screening process to determine the right type of technology for their niche and objective, it will never be a generic translation workflow.
]]> 2018-09-30 21:53:36 UTC https://padlet.com/juan39/hbwu2aiq655c/wish/287390849 How The Funnel Will Look Like (Basic Idea) juan39 https://padlet.com/juan39/hbwu2aiq655c/wish/287390850 Top of the funnel:
  • Inbound and outbound marketing strategy using a case study.
  • The inbound strategy will be around super valuable content, podcasts appearances, campaigns using videos and landing pages for downloading eBooks and informative materials for companies.
  • Possibilities on these regards are virtually limitless. However, social media and advertising should be the main pillars.

Middle of the funnel:

  • The outbound strategy will be cold emailing pitching a case study. Koalingo SHOULD NOT contact startups as these do not have enough budget nor resources to acquire these custom MT-based services.
  • Book meetings and presentations to those warm leads interested in the service.


]]> 2018-09-30 21:53:36 UTC https://padlet.com/juan39/hbwu2aiq655c/wish/287390850 Major Pains (AKA, FEARS) juan39 https://padlet.com/juan39/hbwu2aiq655c/wish/287390851
  • A large array of MT technologies: companies do not know what to choose (daunting testing).
  • FOMO Syndrome: "If I choose Microsoft this year, perhaps Google will come up with an amazing update that strongly beneficiates my specific niche."
  • It might translate things in the wrong way (poor MT implementation): this is the commonest fear and it is well developed in some scientific papers.
  • Compatibility issues with current systems.
  • Lack of trained internal teams on the MT subject.

    • ]]>     2018-09-30 21:53:36 UTC https://padlet.com/juan39/hbwu2aiq655c/wish/287390851