About us
What's happening Stockholm. We are firing up a local MLOps chapter for this amazing city!
The MLOps Community fills the need to share real-world Machine Learning Operations best practices from engineers in the field. While MLOps shares a lot of ground with DevOps, the differences are as big as the similarities. We needed a community laser-focused on solving the unique challenges we deal with every day building production ML pipelines.
We’re in this together. Come learn with us in a community open to everyone. Share knowledge. Ask questions. Get answers.
Upcoming events
1
#40 - Orchestration vs Model Selection
KTH R1, Drottning Kristinas väg 51, 114 28 Stockholm, Sweden, Stockholm, SE***Do note that you will have to sign-up at the Luma event page prior to attending. Do so here: https://luma.com/ltaf6kn7***
All right!!!
Meetup #40 will take place October 21 at KTH Reaktorhallen and focus on "Orchestration vs Model Selection". We're super excited to team up with European consultancy power house CHI Software, Finnish start-up stars Confidential Mind, local heroes Hopsworks and Sweden AI Factory! The program is currently being worked out, so stay tuned for updates!
Event Program
- Doors open at 17:00 CET
- Talks begin at 17:45 CET
- There will be pizzas & drinks
- There may be a moderated Q&A session....Speaker Line-Up TBD...
-
Ihor Tanenkov, AI/ML Architect at CHI Software will will give a talk titled "Sovereign AI in Practice: An Execution Layer for European Organisations". Abstract: Hosting a model locally creates a deployment option. Making it useful for multi-step work takes more engineering: planning, routing, tools, verification and recovery, all inside the organisation's approved boundaries. Ihor will walk through an execution layer above a pool of custom models, open-weight models and deterministic tools, where routing happens at each step of a task and starts with eligibility, not cost. A banking briefing serves as the worked example.
-
Jim Dowling, Co-Founder & CEO of Hopsworks will will give a talk titled "Orchestrating ML systems with Hopsworks ML Software Factory". Abstract: Imagine being a domain expert, but not a programmer, but being able to build and operate software systems like seasoned software engineer. That is the promise of software factories that extend coding agents to manage the full SDLC. A software factory requires (1) a domain for which the software will be developed and (2) a platform that will manage the full SDLC from design to development to operations and maintenance. Software factories work best in domains where you know the requirements and can deterministically verify that the requirements have been met, such as for machine learning (ML) systems. In this talk, we will introduce software factories and describe a software factory for building ML systems using Hopsworks.
-
Akshay Jain, Senior Research Engineer at Confidential Mind will give a talk titled "SampoTron: Closing the Cross-lingual Gap for Nordic languages via Query Routing". Abstract: Nemotron-3-Embed-1B is one of the strongest native-Finnish embedders, but cross-language queries lose more than a third of their recall across all embedder models. Instead of retraining or re-embedding, we added a small query-side router, SampoTron, and closed the gap by 21%, running the whole stack on on-prem. And so, this talk will delve into what one should think when choosing models such as an embedder for their RAG setup and how working across Swedish, Finnish, and English languages can be challenging for off-the-shelf models. Then we will briefly discuss the idea behind ConfidentialMind’s SampoTron model and demonstrate results on how we recover performance for cross-lingual tasks.
-
An ace from Sweden AI Factory.
Looking forward to seeing you there,
/Patrick & the Stockholm MLOps Team
58 attendees-
Past events
46
