AI infra & open-source models #1

​We're hosting our first dstack user group event in Berlin and inviting AI engineers, researchers, and infrastructure experts interested in AI infra and open-source models. Join us for lightning talks, demos, and focused discussions.

​This event designed to bring together individuals deeply involved in the AI space to exchange ideas, share technical insights, and connect.

​Schedule:

• ​6:00pm – Arrival

• ​6:30pm – Lightning talks and demos (Part 1)

• ​7:15pm – Break

• ​7:30pm – Lightning talks and demos (Part 2)

• ​8:15pm – Networking with food & drinks

​Each session features 5-minute talks followed by 5-minute Q&As.

​Speakers:

• ​Andrey from dstack - 'dstack: Beyond K8S and Sturm'

• ​Tim from RunPod - 'Deploy AI apps using natural language'

• ​Scott from Jina AI - 'Jina AI: Feature-rich enterprise-ready models'

• ​Andrey from Qdrant - 'miniCOIL: a new model for Sparse Neural Retrieval'

• ​Tanja from Rasa - 'Keep CALM and Generate Commands: A New Dialogue Understanding Architecture'

• ​Roman from Nixiesearch - 'Serving a 9B embedding model without losing your mind'

• ​David from Lambda - 'Winning the ARC prize 2024'

• ​Szymon from Aleph Alpha - 'How to keep your GPU happy'

• ​Anton from Tokalon.ai - 'Budget-friendly image generation'

• ​Piotr from Aleph Alpha - 'Prompt case vs token by token decodin'

• ​Vedant from Aleph Alpha - 'Steering vectors'

• ​Artem from JetBrains - 'Fast track for data synchronization in AI'

• ​Mish from E2B - 'Building an open-source computer use agent'

• ​Mathis from deepset - 'Building open source agents with Haystack'

• ​Tejas from DataStax - 'Build your own local-first, local-only AI workflow'

• ​Jay from TextCortex - 'Saving your sanity and money with pgvector'

​About dstack:

​dstack is an open-source alternative to Kubernetes and Slurm, simplifying AI development and deployment across clouds and on-prem with support for NVIDIA, AMD, and TPU hardware.

​Thanks to RunPod, Aleph Alpha, and Tensor Ventures for supporting this event.