Fireside Chat with Jarek Kutylowski, Founder & CEO of DeepL | RAISE Summit 2026

DeepL Company Profile & Trajectory

• Founded in 2017 by Jarek Kutsiwowski; approaching nine years of operation as of August 2024.
• Currently serves 100,000 company and organizational customers.
• Built infrastructure from scratch in 2017, including self-sourced GPUs and custom data centers due to the lack of existing AI infrastructure.
• Company mission: Eliminate the necessity of learning languages for professional success, while encouraging cultural and personal language learning.

Technology & Competitive Positioning

• DeepL does not compete on simple translation (e.g., emails), which has become a commodity powered by foundational models.
• Focus has shifted to high-stakes, specialized use cases: technical documents, patent applications, complex legal contracts, and court evidence.
• Competitive advantage stems from specialized training since 2017, distinct model architectures, and optimized training regimes that balance quality, speed (latency), and cost.
• DeepL powers many consumer-facing AI translation features (e.g., YouTube, Airpods) as a backend provider rather than just a direct-to-consumer app.
• While DeepL utilizes Large Language Models (LLMs) internally, it claims superior error rate management compared to general-purpose models to minimize hallucinations, though it acknowledges 100% error-free translation is impossible even with human oversight.
• Emerging focus area: Real-time speech translation to facilitate live business negotiations across languages (e.g., Japanese to South Korean).

Operational Adjustments & Restructuring

• Executed a 20% staff reduction earlier in the year; no rehiring of laid-off employees is planned.
• Restructuring aimed to improve speed, efficiency, and align workflows with AI adoption across internal departments (finance, engineering).
• CEO Kutsiwowski notes that the shift has allowed smaller teams to deliver end-to-end solutions by augmenting core skills with AI.

Market Dynamics & Economics

• Token costs and infrastructure constraints are internal and external concerns, but DeepL is positioned better than application-layer-only software due to proprietary model ownership.
• Corporate customers are increasingly seeking specialized solutions over foundational models to achieve better efficiency per token and lower costs.
• DeepL maintains long-term investor alignment, deferring IPO decisions until the right market conditions emerge, despite recent public market volatility.
• The company views the commoditization of basic translation as a growth feature, as it drives integration of DeepL technology into third-party products.

European Tech Outlook

• Kutsiwowski acknowledges Europe's current "depressing" start position relative to the US but rejects the notion that AI dominance is settled in the US.
• Cites the rapid market shifts (e.g., rise of Anthropic, OpenAI's struggles) as evidence that Europe can still compete through new ideas and agility.
• Argues Europe must work harder to overcome a slight head start disadvantage for US competitors.

Future of the Translation Workforce

• Human translator roles are evolving rather than disappearing; jobs are shifting toward AI program management and workflow integration.
• Technology cannot yet fully replace humans in rare, high-complexity scenarios requiring absolute trust.
• The future market will require more personnel skilled in deploying and managing AI translation tools to handle growing global content volume.

Forward-Looking Vision

• Ultimate goal: Global business operations without language barriers, while maintaining awareness of regulatory, currency, and cultural differences.
• Predicts a utopian future where technology handles linguistic barriers, allowing professionals to focus entirely on strategy and execution.
• Maintains that language learning should remain a voluntary pursuit driven by cultural interest, not economic necessity.