YouTube ADASTRA

Přes 80 % firem už nějakým způsobem nasadilo umělou inteligenci. Zaměstnanci říkají, že jim nástroje šetří čas. Do firemních výsledků se to zatím nepromítá. Petr Zelenka, Chief AI Officer Adastry, byl hostem Ranního brífinku Hospodářských novin. S moderátorem podcastu Štěpánem Svobodou mluvil o tom, proč se produkční nasazení AI daří jen zhruba pětině firem, kde je rozdíl mezi prototypem a reálnou integrací do byznysu, a proč se vyplatí dívat se i mimo velké jazykové modely. Z rozhovoru: - K nasazení AI do produkce se hlásí 18 % firem v Česku a 20 % v Evropě. Poměr na trhu je zhruba jedna ku čtyřem. - Malý prototyp pro malou skupinu je ta snadná část. Těžká část přijde, když má AI fungovat jako součást byznysu. - U generativní AI si zrychlíte generování výstupů, ale ne práci po něm. Někdo je pořád musí přečíst a něco s nimi udělat. - Matematická optimalizace nemá tolik pozornosti, ale návratnost se u ní dá spočítat předem. Kolik kusů navíc vyrobím na stejné lince, o kolik zmenším zásoby, o kolik míň najezdím kilometrů. - Jak nepracovat s jazykovým modelem tak, aby vám jen potvrzoval, že máte brilantní nápad. Zdroj: Ranní brífink Hospodářských novin, 4. září 2026.

9. 9. 2026

A few thousand .sas programs, decades of logic, a model risk team watching. Here’s why we think the smarter way off SAS and onto Databricks combines automation with judgment – not one or the other.

27. 7. 2026

A few hundred streams, some scoring customers every night, most of them older than the people now asking what they do. Here's the three-tier factory we use to move them onto Databricks – deliberately, with evidence attached.

27. 7. 2026

Shingai George, Aviation Consultant specializing in data analytics, AI, and sustainability, shares how modern data platforms, AI-driven decision-making, and shared data ecosystems are helping the aviation industry adapt to a new era of geopolitical volatility, regulatory pressure, and sustainability demands. He explains how airlines are moving beyond decades of cost-and-efficiency optimization toward resilience as a competitive advantage, and how AI is reshaping everything from flight planning and predictive maintenance to passenger experience and emissions management. He also explores why the future of airline success lies in shifting from short-term yield to long-term customer value, using AI to build loyalty and trust, not just revenue. Drawing on experience across customer service, flight operations, route development, flight safety, MRO, and sustainability, he highlights why aviation, one of the world’s most interconnected industries, must move from fragmented optimization toward network-wide intelligence.  This episode will answer:   • How can airlines build resilience into flight planning, fuel forecasting, and operational decisions when geopolitical shocks can reroute entire networks overnight?  • What role does AI play in transforming the passenger journey from reactive service to predictive engagement, without crossing the line between personalization and perceived unfairness?  • How can the industry responsibly share data across airlines, airports, and regulators, and why is federated data sharing key to the future of air traffic management, sustainability, and network resilience? 

9. 9. 2026

České školství se nedá změnit jedním rozhodnutím shora. Podle Tomáše Kaplana z Nadačního fondu Eduzměna je potřeba pracovat přímo v regionech, se školami, řediteli, učiteli, rodiči i zřizovateli. A právě proto začínají hrát zásadní roli data a AI. „Bez pomoci AI nebo moderních technologií si s tím už moc nevíme rady. Museli bychom mít mnohem větší tým,“ říká v novém dílu podcastu Adastry.

6. 7. 2026

• Jak se měří dopad systémové změny ve vzdělávání?  • Proč je těžší pracovat s kvalitativními rozhovory než s tabulkou počtů škol a učitelů?  • A jak může AI (https://adastracorp.com/cs/adastra-ai/) pomoci školám doporučovat vhodné formy podpory podle jejich konkrétních potřeb?

9. 9. 2026

In diesem Vortrag spricht Dr. Patrick Bartsch, Principal Evangelist Automotive bei AWS, über die Anforderungen der Industrie im Bereich Künstliche Intelligenz und darüber, wie Unternehmen mit AWS-Services eigene KI-basierte Anwendungen entwickeln können. Im Fokus steht das Thema Smart Production mit Souveränität: Wie lassen sich moderne KI-Technologien sinnvoll, sicher und praxisnah in industriellen Umgebungen einsetzen? Der Vortrag gibt Einblicke in aktuelle Herausforderungen, mögliche Lösungsansätze und die Rolle von Cloud- und KI-Services für die Produktion der Zukunft.

17. 6. 2026

Chris Peart, Sales Leader at Snowflake Canada, shares how unified data, governed context, and agentic AI are reshaping how enterprises turn information into action. He explains why "context is king" as frontier models become commoditized, how a single AI Data Cloud across AWS, Azure, and GCP removes the brittleness of traditional architectures, and how Snowflake Cortex and Coworker give knowledge workers immediate answers instead of waiting weeks for engineering teams. He also digs into the agentic future and the cultural shift required to win with AI: why "nobody sells anybody anything" and customers buy outcomes, why Canadian enterprises are falling behind global peers by being too cautious, and how the next frontier is autonomous agents negotiating with other agents across organizational boundaries. The episode answers: • Why is your data, not the model you choose, the true competitive differentiator in the agentic era? • What does a governed context layer look like, and why is it the foundation for agents you can trust? • Why are Canadian enterprises falling behind global peers on AI, and what does it take to start swinging?

29. 6. 2026

Justin Rister, Senior Cloud and AI Specialist at Microsoft, explains why leaders should start with business pain points, not technology. He shares how Fabric unifies the analytics stack for teams of all skillsets, why Databricks and Fabric are a better-together story, and how an AI layer on unified data empowers business users to ask questions and get answers without waiting on IT. What does it mean to be a strategic partner instead of a product pusher? How do you remove bottlenecks by letting business users access insights directly? Why should leaders think big, start small, and scale fast?

21. 5. 2026

Brad Freels, Azure Data and AI Specialist at Microsoft, shares how meeting customers where they are, building AI Centers of Excellence, and diving in without waiting for "perfect data" are reshaping how enterprises unlock value from cloud, data, and AI. He explains why the biggest barrier to AI isn't technology but fear and paralysis, how partners compress time to value by bridging Microsoft best practices to each customer's unique environment, and why executive-sponsored Centers of Excellence lower the temperature on change management while creating guardrails that prevent shadow AI and governance nightmares. He also shows where to start now, such as standing up a 60-day Fabric trial, rebuilding an existing report, and letting AI itself help curate the messy data you already have, because your competitors aren't waiting and neither should you.  • What does it really take to move from fear of AI to a culture that leans into it?   • How can you start small with the data and tools you already have and still build toward enterprise-wide transformation?   • Which guardrails, sponsorship, and partner relationships let you accelerate AI adoption without losing control? 

29. 6. 2026

Brandon Ash, Director of Solution Engineering at Microsoft, shares how enterprise leaders can navigate the AI era by building governed, unified data foundations. He explains why the semantic layer is more critical than ever, how OneLake enables a "skunkworks with guardrails" approach, and why customers who embrace partners go further, faster.  This podcast will answer:  • What's holding executives back from AI, and how does semantic debt compound the problem?  • How does Fabric bridge the gap between analytics and operational decision-making?  • Why should you think of AI agents as just another employee to onboard? 

29. 6. 2026

Justin Rister, Senior Cloud and AI Specialist at Microsoft, explains why leaders should start with business pain points, not technology. He shares how Fabric unifies the analytics stack for teams of all skillsets, why Databricks and Fabric are a better-together story, and how an AI layer on unified data empowers business users to ask questions and get answers without waiting on IT.  • What does it mean to be a strategic partner instead of a product pusher?   • How do you remove bottlenecks by letting business users access insights directly?  • Why should leaders think big, start small, and scale fast? 

29. 6. 2026

Tamer Farag, Global Fabric Partner Lead at Microsoft, shares how the fastest-growing analytics platform in the world is helping 31,000 customers unify fragmented data estates and unlock AI value. He highlights why you don't need to move your data to govern it, how mirroring is offered free to accelerate adoption, and what makes partners like Adastra critical to scaling Fabric globally.  • What does it take to connect AI to your data without a massive migration project?   • How is Fabric enabling customers to move from static reports to asking questions directly to their data?  • Which trends, from real-time intelligence to chat with your data, are driving customer demand in 2026? 

29. 6. 2026

Tamer Farag, Global Fabric Partner Lead at Microsoft, shares how the fastest-growing analytics platform in the world is helping 31,000 customers unify fragmented data estates and unlock AI value. He highlights why you don't need to move your data to govern it, how mirroring is offered free to accelerate adoption, and what makes partners like Adastra critical to scaling Fabric globally. What does it take to connect AI to your data without a massive migration project? How is Fabric enabling customers to move from static reports to asking questions directly to their data? Which trends, from real-time intelligence to chat with your data, are driving customer demand in 2026?

28. 4. 2026

Jon Steffey, Senior Director of Enterprise Software and Analytics at Tolmar, shares how a unified data platform, risk‑aware governance, and AI‑enabled workflows are helping a mid‑sized pharma manufacturer modernize without losing sight of quality. He explains how Tolmar is digitizing manual processes, breaking down post‑M&A data silos, and using Microsoft Fabric to move leaders from data overload and gut feel to more empirical decisions. Drawing on experience in aerospace, medical devices, and pharma, he shows why data is foundational to process control in regulated manufacturing and how to match the rigor of controls to real product and patient risk. How do you modernize a legacy, siloed pharma data landscape while focusing on fundamentals like ingestion, transformation, and governance instead of chasing a single “killer” use case? What changes when leaders move from fragmented reports to a unified view of end‑to‑end manufacturing and quality data? How can pharma organizations tune development and validation rigor to the impact and risk of each use case, from low‑risk UI changes to high‑impact, AI‑influenced therapies?

20. 4. 2026