Lena Fuhrimann_

A practical walkthrough of running LLMs locally and on-prem - covering model selection, inference performance, and keeping sensitive data out of third-party hands. Given as an internal Tech Talk at Karakun.

Why running your own LLM makes sense - both as a business decision and as a matter of data sovereignty. A practical look at self-hosted models, the tradeoffs versus managed APIs, and how to get started without a data center.

A panel on digital sovereignty at Cloud Native Zürich 2026 - exploring what it means for organizations to own their infrastructure, data, and AI stack in an era of hyperscaler dependency.

A practical walkthrough of running LLMs locally and on-prem - covering model selection, inference performance, and keeping sensitive data out of third-party hands.

Netzwoche covers the second Open Source AI Conference in Bern - featuring speakers on open Swiss language models, AI ethics, and practical tools for running your own LLM without depending on big tech.

A 5-minute, 20-slide Ignite take on adopting AI where it actually adds value - without creating chaos or risk.

Clearing up the confusion around serverless - why it's a spectrum, not just FaaS, and why the real saving is engineering time, not the cloud bill.

How bespinian grew to serve 100+ clients - the founding story, building a team, and lessons from running a cloud-native consultancy in Switzerland.

From the December 2025 Kubernetes Community Days Suisse Romande at CERN - how to host LLMs on Kubernetes to keep data under your control.

Panel discussion at Kubernetes Community Days Suisse Romande - exploring whether Kubernetes is the right platform for AI workloads, and how AI might change the way we operate Kubernetes.

Guest lecture on cloud-native migration challenges and serverless best practices for the Cloud Development module at OST.

Building strong cloud applications around four ideas - Tolerance, Fortitude, Resilience and Adaptability.

Full house at bespinian HQ for DevOps Meetup Bern - a talk on where AI adoption actually adds value in engineering teams, without the hype.

Deploying and managing large language models on Kubernetes - Ollama for simple setups, the vLLM Production Stack for high-traffic production.

Avoiding common serverless pitfalls - logging, tracing, cold starts, security and deployment strategies.

What NixOS is, how it works, and the genuinely nice things you can do once your whole machine is declarative and reproducible.

A decade of building and running cloud-native and serverless systems - the patterns that held up, the mistakes worth avoiding, and what the industry got right and wrong along the way.

Playing through the ten levels of writing and deploying serverless apps - the challenges teams hit, and the best practices that get you past them.

Live drawing instead of slides - juggling a microphone and drawing pen at the same time at my first Swiss Cyber Security Days.

Busting serverless cost myths with real-world examples - how proactive cost monitoring, ARM architectures, and tools like AWS Lambda Power Tuning keep serverless from becoming a cost trap.

A deep dive into how AWS Lambda and Fargate work under the hood, including a live demo of Firecracker - the microVM technology powering them.

How Firecracker launches microVMs in under 100ms - and what that means for serverless Rust on AWS Lambda.

Efficient large-scale serverless deployment strategies using AWS Lambda Powertools at the relaunched DevOps Basel community meetup.

Best practices for building and operating serverless applications in TypeScript, with open-source workshop resources.

Building resilient and reliable serverless applications on AWS - covering optimization strategies for performance and cost at the Cloud Natives Zurich meetup.

Two contributions at SLACon Berlin: a talk busting serverless cost myths and an advanced serverless workshop.

A practical workflow for debugging workloads in Kubernetes without losing your mind - or your weekend.

My declarative, reproducible machine setup in Nix - the config behind the 'how my laptop became declarative' talk.

Building beautiful terminal UIs in Go using the charm.sh library suite.

Playing through the ten levels of writing and deploying serverless apps - the challenges teams hit, and the best practices that get you past them.

Technologies for defining service contracts - OpenAPI, GraphQL and gRPC - and keeping specification and implementation in sync.

Combining Git and Infrastructure as Code to implement GitOps for both apps and infra - Argo CD for Kubernetes, Terraform for infrastructure.

A profile and interview about my path into cloud software architecture, co-founding bespinian, and the intersection of technical expertise and creativity.

IaC principles and declarative infrastructure with Terraform, Packer and CloudFormation - versioning, testing and automation.

Git best practices for teams - commit messages, Gitflow vs GitHub Flow, pull requests and hooks to enforce conventions.

Hands-on serverless workshop at Swiss Cloud Native Day, co-facilitated with Gabriel Koch and Mathis Kretz.

A concise step-by-step guide to installing Arch Linux on UEFI systems with LUKS full-disk encryption.

Moving around Vim efficiently - movement keys, search techniques and a few plugins that boost your editing speed.

Deploying Zenbot, an open-source cryptocurrency trading bot, to Cloud Foundry to automate trading strategies on platforms like GDAX.

Deploying WordPress on Cloud Foundry with externalized S3 file storage to handle media uploads in a cloud-native way.

A private Docker Registry on Cloud Foundry using S3 for storage and Redis for caching to securely manage your container images.

Deploying Ghost, a modern Node.js blogging platform, on Cloud Foundry with MariaDB and optional S3 storage for media files.

Using Docker to spin up and tear down backend services for local development - keeping your machine clean and your workflows efficient.

A fun multiplayer game: players swipe on their phones to launch balls at targets on a shared big screen, with a live scoreboard.

Creating and managing S3 buckets on Cloud Foundry using s3cmd or the open-source S3 Manager application.

Zero-downtime updates and safe rollbacks on Cloud Foundry using a simple shell script.

A web GUI written in Go to manage S3 buckets from any provider - list, upload, download and delete objects. Runs as a tiny container, deploys anywhere.

Deploying Piwik, an open-source analytics platform, on Cloud Foundry as a self-hosted alternative to Google Analytics.

Configuring static front-end apps on Cloud Foundry via environment variables exposed through an nginx HTTP endpoint.

Using Route Services to add authentication, rate limiting or logging at the route level before requests reach your app.

Angular services that consume Feathers WebSocket APIs using RxJS Observables for real-time, reactive data streams.

Deploying Strider, an open-source CI/CD server, on Cloud Foundry to test, build and deploy your apps on every git push.