Join Latvia DevOps & AI Communities’ 31st event on May 5 at Workland. Doors open at 18:30, talks start at 19:00.
Janis is a product leader with 17+ years in B2B SaaS, founded Mailigen (acquired by Pipedrive), scaled product at Pipedrive during its unicorn growth phase, and most recently built the product organization from zero at Aerones, a deep-tech robotics company in the wind energy space. He is passionate about AI-first product development and rethinking how product teams are structured in the age of AI dev tools.
Talk name: From Zero to AI-First: How We Reinvented Product Roles at a Deep-Tech Startup.
Most companies talk about adopting AI tools. Few talk honestly about what happens when you try to rewire an entire organization around them — the resistance, the role confusion, the wins, and the spectacular misfires. At Aerones, Janis built the product org from scratch for a company that maintains wind turbines with robots. Along the way, two new roles were introduced — Product Builder and Citizen Builder — designed to push AI-assisted development beyond the engineering team and into the hands of product managers and domain experts.
This talk covers the full journey: why these roles were created, how the enablement program was structured, what tooling and workflows actually worked, where things went wrong, and what would be done differently. It is a practitioner’s story — no vendor pitches, no hype, just lessons from doing it in a complex, hardware-adjacent environment. You will walk away with a practical framework for thinking about AI-assisted development roles, and an honest picture of what organizational transformation actually looks like when the PowerPoint slides meet reality.
Māris Svirksts is a Lead Systems Engineer at EPAM Systems with 15+ years of experience in tech, ranging from boutique hotel booking platforms to cloud infrastructure for global corporations, and occasionally explains to stakeholders why “just make it work” is not a technical specification. He speaks Latvian, English, and sometimes Bash. The AI writing this bio was made by one of his employers’ clients, which he finds either impressive or slightly unsettling.
Talk name: THE DARK FACTORY / From Loom-Smashing to Prompt Engineering.
From the Luddite riots against the power loom to today’s robotic production lines, every automation wave has fundamentally changed how work is organized — and the software industry is living through the next one. While tools like Cursor and Claude have already introduced AI-assisted coding and autonomous agents, an LLM-driven dark factory for the software development lifecycle takes this further, stringing together the full cycle of planning, development, testing, and delivery with minimal human intervention.
We will also have a Lightning Talk by:
Vija Kalniņa is a lead researcher on the Business AI Ethics research initiative based at the Edmond & Lily Safra Center for Ethics at Harvard University. She is affiliated with AI Master Lab, an AI consulting firm based in Latvia, and teaches AI ethics to different audiences, including judges and law students. She is a Hubert H. Humphrey (Fulbright) alumna at Boston University’s Questrom School of Business (2024/2025). She has 15 years of experience in the legal field across the public and private sectors, including her most recent role as a judge at the Court of Economic Affairs of the Republic of Latvia, where she worked with white-collar crime cases and commercial disputes. She holds a PhD in European Union Law, is a published author on EU Law, and a freelance interviewer for Latvia’s leading legal journal, Jurista Vārds.
This lightning talk uses AI chatbot doom loops – situations in which users are trapped in repetitive, unresolvable interactions – as a lens for learning how to examine AI incidents. By focusing on these failure patterns, the talk highlights how harm can emerge when AI systems fail. I will use a real-world example – a U.S.-based online payment and financial technology provider that relies on AI-powered chatbots for client services, and the incidents it experienced – to illustrate how issues such as insufficient human oversight, poor escalation design, and model or system limitations can lead to reputational and financial harm for companies, as well as frustration, lost time, reduced access to services, or even financial consequences for clients. The goal is to spark discussion and help the audience strengthen their ability to recognize AI-related risks and negative consequences as they arise, because understanding AI incidents is a necessary first step toward preventing them..
This time, the event is sponsored by:
- EPAM as Event Sponsor
EPAM is a global leader in AI transformation engineering and integrated consulting, helping Forbes Global 2000 companies and ambitious startups to become AI-native enterprises, driving measurable value from innovation and digital investments.
We are proud to be recognized by Forbes, Glassdoor and Great Place to Work as a Most Loved Workplace around the world. Learn more at www.epam.com and follow us on LinkedIn.
- Workland Telegraph as Event Partner
Inspiring private offices and coworking spaces that support your productivity & creativity.
If your plans change, please cancel your registration to free up your spot for others.
Also, check out our Slack channel Techies of Baltics.
As always, you are welcome to check out our YouTube channel, where we have previous meetup video recordings available.
Your presence and active participation make our community vibrant and inspiring. To secure your spot, RSVP on our Meetup.com page.
See you there!
*Please note that photographs and footage will be taken throughout this event without a warning. These can be used by the Community for marketing and publicity in our publications, on our website and in social media, or any third-party publication.
