Sign in

Luke Marsden

@lmarsden.bsky.social
2.6K followers 4.5K following 84 posts

Hacker & entrepreneur. Founder helix.ml, private GenAI stack, getting business value out of local open source LLMs

PostsRepliesMedia
Luke Marsden @lmarsden.bsky.social · 02/06/2026
Yep, it pairs nicely with CDD - conference driven development 🤣
000
Luke Marsden @lmarsden.bsky.social · 02/06/2026
Gonna test this out on the train today 💪💪💪
110
Luke Marsden @lmarsden.bsky.social · 02/06/2026
Wowee you are amazing!
010
Luke Marsden @lmarsden.bsky.social · 01/08/2025
Bootstrapped a Private GenAI Startup to $1M revenue, AMA blog.helix.ml/p/bootstrapp...
blog.helix.ml
Bootstrapped Private GenAI Startup Hits $1M Annual Revenue, Launches Helix 2.0
The people behind the story, how agentic AI is changing and why we don't want a sales call with you
041
Luke Marsden @lmarsden.bsky.social · 03/04/2025
Thanks Hannah! 💖
050
Luke Marsden @lmarsden.bsky.social · 03/04/2025
Good morning #KubeCon! Come find me if you want to talk running LLM systems like vision RAG in production on Kubernetes today. Email luke@helix.ml or reply here!
050
Luke Marsden @lmarsden.bsky.social · 02/04/2025
Phil setting up a vLLM vision model provider on helix on his nerd phone on the Elizabeth line #kubecon
050
Luke Marsden @lmarsden.bsky.social · 02/04/2025
I mean, yes please 😄
000
Luke Marsden @lmarsden.bsky.social · 02/04/2025
Catch my talk with Priya Samuel on running LLMs in production on k8s today at #KubeCon at S10A at 3:15pm ✨
082
Luke Marsden @lmarsden.bsky.social · 02/04/2025
Trying (but failing) to be as colourful as @hannahfoxwell.net today!
110
Reposted by Luke Marsden
HelixML @helixml.bsky.social · 01/04/2025
Oh hello there BlueSky! We've arrived here just in time to share all the highlights of #KubeCon #CloudNativeCon in sunny London! Don't miss @hannahfoxwell.net and @lmarsden.bsky.social at 15:20 today talking about Platform Engineering and Developer Experience For Your On Prem LLM
052
Reposted by Luke Marsden
Jennifer Riggins @jkriggins.bsky.social · 20/02/2025
I looked to define a term that’s being thrown around — #AI-native developer — with the help of @guypo.com @lmarsden.bsky.social and Thoughtworks’ Mike Mason so you can prepare your software teams for the near future. thenewstack.io/what-is-an-a... What’s your definition? Only on @thenewstack.io
thenewstack.io
What Is an AI Native Developer?
We explain what an AI native developer is, how the role will evolve, and why spec-driven AI development is driving this trend.
142
Luke Marsden @lmarsden.bsky.social · 20/02/2025
Based on my experience of riding the crazy horse that is cursor agent mode, spec-driven development is much needed!
020
Luke Marsden @lmarsden.bsky.social · 30/01/2025
blog.helix.ml/p/you-should...
blog.helix.ml
You should run local models: run Deepseek-R1 on Helix 1.5.0 now
Chinese Hedge Fund crashes US stock market by releasing Deepseek-R1, and what this means for your enterprise GenAI strategy
010
Luke Marsden @lmarsden.bsky.social · 30/01/2025
Deepseek are clever fuckers. I wrote this about how Deepseek is pushing decision makers in large financial institutions to seriously consider running their own models instead of calling out to Microsoft, Amazon & Google Link below 👇
330
Reposted by Luke Marsden
OpenUK @openuk.bsky.social · 15/01/2025
Speaker @lmarsden.bsky.social of MLOps Consulting will be in our SOOCon25 AI Openness track! 🎤 Join to hear about Production-Ready LLMs on Kubernetes: Patterns, Pitfalls, and Performance.💡https://buff.ly/3HjeQxq #opensource #opensourceai #ai #stateofopencon #soocon25 #opensourcelondon #openuk
062
Reposted by Luke Marsden
Martin Paul Eve @eve.gd · 12/01/2025
I enjoyed Stories of Your Life and Others so much that I'm moving straight to Chiang's other collection: Exhalation.
The cover of Ted Chiang's Exhalation. It shows a pair of lungs made up by plants and gears/cogs.
2102
Luke Marsden @lmarsden.bsky.social · 08/01/2025
MS Teams in the UK this morning is the most grim experience imaginable, tons of jitter and bandwidth seems fucked - like trying to do a video call over a highly contended LTE connection - my network connection is fine (per mtr) - is there an outage in some MS data center somewhere?
020
Luke Marsden @lmarsden.bsky.social · 06/01/2025
We shipped a lot over Christmas and I came here to release it: github.com/helixml/heli... - You can now drag'n'drop files directly into knowledge for Helix apps (rather than having to go via the filestore). - Initial support for MCP (more on this coming soon!)
github.com
Release 1.4.12 - apps drag'n'drop filestore, initial MCP support · helixml/helix
What's Changed make frontend work better with filestore knowledge by @nessie993 in #668 Model Context Protocol (MCP) support by @nessie993 in #680 security/refactoring/quality changes: Revive vo...
030
Luke Marsden @lmarsden.bsky.social · 03/01/2025
Cancelled my X subscription over all the stupid political interference
160
Luke Marsden @lmarsden.bsky.social · 20/12/2024
Thank you to @dciangot.com for doing the heavy lifting getting HelixML GPU runners running on Slurm HPC infra to take advantage of hundreds of thousands of GPUs running on Slurm infrastructure and transform them into multi-tenant GenAI systems! Read all about it here: blog.helix.ml/p/running-ge...
blog.helix.ml
Running GenAI on Supercomputers: Bridging HPC and Modern AI Infrastructure
Transform Your Supercomputer into a Private OpenAI: A Look at How Modern HPC Infrastructure Can Power Enterprise AI
081
Luke Marsden @lmarsden.bsky.social · 20/12/2024
This is so awesome! Imagine being able to combine the worlds of cloud-native scalable web services for LLMOps with the raw power of Slurm-powered supercomputers with some of the biggest compute power, networking and GPUs! Check out the writeup here: blog.helix.ml/p/running-ge...
blog.helix.ml
Running GenAI on Supercomputers: Bridging HPC and Modern AI Infrastructure
Transform Your Supercomputer into a Private OpenAI: A Look at How Modern HPC Infrastructure Can Power Enterprise AI
040
Luke Marsden @lmarsden.bsky.social · 19/12/2024
So @cybernetist.com got gptscript working with llama3.3:70b! Check out the detailed writeup here: blog.helix.ml/p/gptscript-...
blog.helix.ml
GPTScript Helix Apps: For Fun and Profit
Making GPTScript Shine with Open Source LLMs: How Llama 3's 70B Model Finally Makes Natural Language Programming Reliable
053
Luke Marsden @lmarsden.bsky.social · 13/12/2024
Got helix.ml running GenAI on a supercomputer..
helix.ml
Private GenAI Platform – HelixML
Build, test and deploy GenAI apps. Connect knowledge and business applications via APIs. Deploy in your cloud or use ours
030
Luke Marsden @lmarsden.bsky.social · 11/12/2024
You're a dope llama hacker (ha ha try finding a gif for that)
000
Luke Marsden @lmarsden.bsky.social · 11/12/2024
Fucking dope helix.ml release from @cybernetist.com if you like function calling in open source LLMs and @GPTScript_ai github.com/helixml/heli... (Excuse my French)
github.com
Release 1.4.9 - support function calling in helix runners, add gptscript runner · helixml/helix
What's Changed Add gptscript_runner service to helix stack by @milosgajdos in #635 Bump ollama and update LLM inference with Tool calling by @milosgajdos in #637 Full Changelog: 1.4.8...1.4.9
010
Luke Marsden @lmarsden.bsky.social · 04/12/2024
Here's the post with the youtube video walkthrough of the workshop in it: blog.helix.ml/p/building-r...
blog.helix.ml
Building Reliable GenAI Applications: A Hands-on Testing & CI Workshop
Recap and a walkthrough video of the Testing & CI for GenAI Workshop we ran yesterday. Join the next one!
010
Luke Marsden @lmarsden.bsky.social · 04/12/2024
Test Driven Development (TDD) for your LLMs? Yes please, more of that please! Back to basics - write a test, see the test fail, improve the prompt, see it pass, check it in - just like you would with any other code 😄
231
Reposted by Luke Marsden
Derrick Cogburn @derrickcogburn.bsky.social · 28/11/2024
Luke, this post, including the video, was very helpful. I loved the conference talk. I took my graduate students through a somewhat similar exercise last week for their last lab of the semester. Well done. Thank you.
021
Reposted by Luke Marsden
Derrick Cogburn @derrickcogburn.bsky.social · 28/11/2024
This is a wonderful post by @lmarsden.bsky.social Great presentation and video as well.
021
Luke Marsden @lmarsden.bsky.social · 28/11/2024
ICYMI, here's my 2-part blog series on how we can all be AI engineers with open source models – and how to apply testing best practices to LLM apps 1. blog.helix.ml/p/we-can-all... 2. blog.helix.ml/p/from-click...
blog.helix.ml
We can all be AI engineers – and we can do it with open source models
The barriers to AI engineering are crumbling fast
140
Luke Marsden @lmarsden.bsky.social · 24/11/2024
The reference architecture above uses option (a) for development and the whole stack runs on kind on your laptop And you're right without GPUs you lose fine tuning and also image models. But you can still do knowledge, integrating the LLM with APIs via OpenAPI spec and tests
010
Luke Marsden @lmarsden.bsky.social · 24/11/2024
Yes, if you don't want to run your own GPUs, you can a) use an external LLM provider like togetherai (but then you lose the benefits of total data privacy within the cluster) or b) use ollama with CPU inference, but it's fairly slow unless you have lots of cores and use a smaller model
110
Luke Marsden @lmarsden.bsky.social · 23/11/2024
@llm360.bsky.social please consider adding me to the open source LLMs starter pack ☺️
110
Luke Marsden @lmarsden.bsky.social · 23/11/2024
So how can we _actually_ do AI on Kubernetes? Here's my humble proposal: LLM apps and tests as data (CRDs). Deeply integrated into the existing CI/CD pipeline e.g GitLab and Flux. Self hosted LLMs on GPUs if you want to. Runnable code: github.com/helixml/gena...
github.com
GitHub - helixml/genai-cicd-ref: Cloud-Native GenAI Reference Architecture
Cloud-Native GenAI Reference Architecture. Contribute to helixml/genai-cicd-ref development by creating an account on GitHub.
170
Luke Marsden @lmarsden.bsky.social · 23/11/2024
Hey, nice to meet you! Bluesky feels really fresh, no?
000
Luke Marsden @lmarsden.bsky.social · 23/11/2024
Greetings!
010
Luke Marsden @lmarsden.bsky.social · 23/11/2024
xkcd.com/378/ 😄
xkcd.com
Real Programmers
010
Luke Marsden @lmarsden.bsky.social · 22/11/2024
Just followed all 1400 people who followed me. If you see this, like this post to say hi! 🦋💚🦋
3280
Luke Marsden @lmarsden.bsky.social · 21/11/2024
I wrote a quick summary of the latest MLOps Community podcast: blog.helix.ml/p/from-click...
blog.helix.ml
From ClickOps to GitOps: The Evolution of AI App Development
How we bridge the gap between no-code AI prototypes and production-ready applications
032
Luke Marsden @lmarsden.bsky.social · 20/11/2024
And still, building like this is hard and not without its challenges. Thanks to Chris Sterry and Phil Winder for taking the hard road with me, and Matt Barker, Tamao Nakahara, Dan Baker, and the other people behind the scenes, you know who you are, thank you 😉
000
Luke Marsden @lmarsden.bsky.social · 20/11/2024
National data regulation, repatriation of data, and return to the private cloud is combining with the demand to stay competitive to drive interesting pipeline for our Private GenAI business
100
Luke Marsden @lmarsden.bsky.social · 20/11/2024
One year ago, I promised to build helix.ml in the open, so here's an update on our commercial progress: blog.helix.ml/p/revenue-an...
blog.helix.ml
Building in the open: Revenue and Pipeline for Private GenAI
National data regulation, repatriation of data, and return to the private cloud is combining with the demand to stay competitive to drive interesting pipeline for our Private GenAI business
120
Reposted by Luke Marsden
Simon Willison @simonwillison.net · 17/11/2024
Anyone got a good system in place for running automated evals against your LLM-driven applications, to help test if tweaks to prompts get better results? I'd love to hear about approaches that are working for people on this
15876
Luke Marsden @lmarsden.bsky.social · 18/11/2024
We put together a helix test command that does this, letting you test and version API and knowledge integrations in the same file: github.com/helixml/heli... We are using it to develop e.g Jira integration, demo here youtu.be/XR2DicY1zEM?...
github.com
helix/examples/test at main · helixml/helix
Multi-node production GenAI stack. Run the best of open source AI easily on your own servers. Easily add knowledge from documents and scrape websites. Create your own AI by fine-tuning open source ...
000
Luke Marsden @lmarsden.bsky.social · 17/11/2024
Now that loads more people are here I guess I should run rerun this thing, right?
021
Luke Marsden @lmarsden.bsky.social · 17/11/2024
I'm interested in using aispec.org app specs to generate synthetic training data that after fine tuning makes a small model that's really good at, like, calling the Salesforce API or some other specific task. Wdyt?
aispec.org
AI Spec
011
Luke Marsden @lmarsden.bsky.social · 17/11/2024
Hey cool this thing is here now 🤣
021
Luke Marsden @lmarsden.bsky.social · 17/11/2024
Sign me up! The wheels on the bus go round and round... (literally what I'm hearing right now)
020
Luke Marsden @lmarsden.bsky.social · 17/11/2024
Please consider adding me 😊
120