Generative AI Summit Presentation
Logotype for Straker Limited

Straker (STG) Generative AI Summit Presentation summary

Event summary combining transcript, slides, and related documents.

Logotype for Straker Limited

Generative AI Summit Presentation summary

8 Oct, 2025

Generative AI disruption and industry challenges

  • Generative AI (genAI) is perceived as disruptive, raising concerns about its impact on traditional translation and localisation business models.

  • GenAI excels in major languages but struggles with less common ones, and is prone to hallucinations and factual errors.

  • Human insights remain critical, as genAI lacks the ability to provide nuanced, context-driven judgment.

  • Industry interest is shifting from content generation to quality evaluation and verification workflows.

  • Cost efficiency and process simplification are key drivers for adopting AI-enabled solutions.

AI-powered solutions and workflow innovations

  • Developed a genAI fact-checker and quality estimation engine using proprietary language data and models.

  • Introduced Automated Post Editing (APE) where multiple LLMs collaborate to refine translations, with human verification as a final step.

  • Orchestrate AI workflow builder allows customers to configure translation processes based on budget and quality needs.

  • The Verify AI Agent process flow integrates LLM content generation, quality evaluation, automated post-editing, and human review.

  • Six key components: LLM content generation, SLM quality evaluation, multi-LLM refinement, human verification, configuration engine, and token-based payment.

Business model transformation and cost structure

  • Traditional model: machine translation plus human review, charged per word or hour, resulting in higher total costs.

  • New model: token-based charging for LLM content generation, quality evaluation, and automated post-editing, with human verification as needed.

  • Example cost comparison shows significant reduction in total cost using the AI-driven workflow.

  • Pricing disruption accompanies technological disruption, moving from per-word to token/character-based billing.

  • Customers can tailor workflows to balance cost and quality, triggering human review only when AI quality falls below thresholds.

Partial view of Summaries dataset, powered by Quartr API
AI can get things wrong. Verify important information.
All investor relations material. One API.
Learn more