UTICamp-2020. Is translation automation a myth or a reality?
Download as PPTX, PDF0 likes17 views
The document discusses the complexities of translation automation, emphasizing the roles of various stakeholders like LSPs and project managers in workflow implementation. It highlights three levels of automation—transaction, project management, and linguistic—and emphasizes the importance of understanding APIs and their use in translating functionalities. A successful automation-focused project manager should prioritize process efficiency, collaboration, and data-driven decision-making.
UTICamp-2020. Is translation automation a myth or a reality?
- 1. Translation automation: myth or reality? István Lengyel, 25 July 2020, UTICamp
- 2. GOALS • Understanding automation complexity • Demistifying APIs • Looking at a business case • Identifying the profile of automation-focused PMs TERMINOLOGY • LSP, MLV, SLV, Automator • API, internal API, web scraping
- 3. Enterprise tries to solve a business problem Superagencies (large MLVs) consult on workflows and implement a solution SLVs hire and train linguistic talent Freelancers or in-house linguists do the work The freedom of choice pyramid
- 5. Enterprise Superagencies SLVs Freelancers or in-house linguists Lowest budget, highest requirements
- 6. Job creation Job acceptance Specification/discussion (quote/time) Vendor assignment Package extraction / file download-upload Job delivery and closing Critical moments in a TMS project
- 7. Three levels of translation automation: • Transaction automation (with client) • Project management automation • Linguistic automation / productivity
- 9. Same problems, different tools. Sounds normal? Why are there so many tools out there?
- 11. What do you need to know about APIs? • APIs allow programs to talk with each other • Not every API is public • REST vs SOAP • Use cases in public APIs • Authentication and availability • Postman – get to know • API and business logic – don’t assume developers know what to do
- 12. DEMO – text classification and keyword extraction Postman Integromat
- 13. The business case for automation • Customer value • Saving time – what’s really taking the time? How many simultaneous projects can you take? • Automation and exceptions – How many simultaneous projects can you take? (of a kind) Costs: start at processes, not at automation!
- 14. Automation-savvy project / program / account manager • Similar to a good manager: feels happy if they can move on – use this for incentivizing • KPIs: number of projects running + time spent on research and improvements • Can make informed decisions and can responsibly spend a small budget – ownership of their tasks • Process mantra and exceptions are always done for a reason • Isn’t afraid to involve anybody in a process, including the customer • Is focused on understanding reasons, rather than realities • Rewards collaboration, is ready to understand and address concerns
- 15. Is a good translation automation project manager just the same as a good project manager from any other area?
- 16. Hope this is useful: https://belazy.cat/preparing-translation-company-automation/ And you should read this too: https://www.smartcat.ai/blog/translation-project-management-on- autopilot/ And remember: technology is just a slave to process.
- 17. Thank you!