Hacker News: stevekaram

New comment by stevekaram in "Our $445M Series D"

stevekaram — Fri, 09 Oct 2026 22:08:15 +0000

Before I joined Oxide I did a little fan-boying (as one does) and put together home lab components as best I could. You _could_ tie 3+ mini-PCs together and wrangle omicron into some semblance of working that way. But it would not be pretty.

However! You absolutely can use non-simulated Omicron: https://github.com/oxidecomputer/omicron/blob/main/docs/how-...

You'll get a single node control plane that's mostly workable, and be able to spin up VMs all day with it. That's what I did. Additionally, if you want to play with microcontrollers you can install Hubris onto STM32H753ZI dev boards! I have one connected to my TrueNAS server over UART and USBTTY for sensors, power management, and console.

New comment by stevekaram in "Our $445M Series D"

stevekaram — Fri, 09 Oct 2026 16:01:38 +0000

Enterprise software and hardware, built from the ground up. Definitely not just a VMware alternative, though the end product is virtual instances of course.

We're building racks over here! And all our software is open source.

New comment by stevekaram in "Our $445M Series D"

stevekaram — Fri, 09 Oct 2026 16:00:27 +0000

Not saying we are or aren't looking for that kind of candidate, but I _will_ say that we're very keen on the motivation to join Oxide at all, and the role you're applying for specifically.

New comment by stevekaram in "Our $445M Series D"

stevekaram — Fri, 09 Oct 2026 15:59:22 +0000

Sans AI is definitely the way to go if you plan to submit!

As for the materials, they play a part in the conversation stage, and even if you get hired and interview others hoping to join. Every interviewer reads the materials of the candidate, and we send the interviewer's original materials to the candidate. It's not nearly as tedious as some of these comments are making it out to be (and believe me, I've been through the "ten elevator pitches" gauntlet in my career).

tbh, I really enjoyed filling out my Oxide materials. They got me thinking and taking stock of my wins and motivations, not just the crap we throw on resumes to attract the eye.

New comment by stevekaram in "LLM Stack for 2024 – Initial Survey"

stevekaram — Mon, 19 Feb 2024 19:40:37 +0000

I'd also add that a huge use for LLMs and vectors in the enterprise is to build queries against production data. Keeping the vector DB external to your RDBMS or other production data store is a unique chance to amplify performance without excess latching and other performance hits against the same database you count on for day to day business. Like external super smart indexes.

New comment by stevekaram in "LLM Stack for 2024 – Initial Survey"

stevekaram — Mon, 19 Feb 2024 19:35:28 +0000

I think one of the biggest struggles small startups and practitioners are facing is lack of a good option between "I wonder if this works" and "ready for prime time." Running locally is an option with consumer hardware but is cost prohibitive for a team. Cloud providers are full of complications and hidden costs. Tools like Friendli and Bento are good but ambiguous on costs and get difficult to price end-to-end once you need the full stack of options. Hugging Face inference endpoints and other tools still seem like the best option around along with cloud DBs like Zilliz.

That said, it's no wonder people just pay extra for the simplicity of a slightly smarter endpoint like OpenAI. Sure, over time the costs are insane and you lack any flexibility to create a truly targeted solution, but it feels like an all-in-one easy fix.