Technology engineering & research
We build difficult things.
North Star Software designs, builds, secures, and operates AI, software, and infrastructure for organizations that need it to work — and tells you plainly when something shouldn't be built at all.
- Senior engineersYou work with the people doing the building.
- You own itCode, accounts, and infrastructure stay yours.
- Straight answersIncluding when the answer is not to build.
- 24-hour repliesOn weekdays, from Minneapolis, MN.
Services
Six areas, one senior team
You don't need to know whether you need an AI consultant, a software developer, a systems engineer, a security firm, or a research team. That's our problem.
- 01
Artificial Intelligence
Private AI deployments, local inference, AI agents, automation, knowledge systems, integrations, enterprise AI architecture, and custom AI solutions.
- 02
Software Engineering
Web applications, APIs, SaaS platforms, internal tools, business systems, automation, integrations, data pipelines, and custom software.
- 03
Infrastructure
Servers, cloud systems, networking, containers, databases, hosting, CI/CD, monitoring, backups, deployment and operations.
- 04
Cybersecurity
Security engineering, hardening, isolation, access control, secure architecture, cybersecurity R&D and technical investigations.
- 05
Web / SaaS
Websites, hosting, SaaS development, domains, integrations and practical technology for small and medium-sized organizations.
- 06
Research & Development
Experiments, prototypes, emerging technology, technical research, proof-of-concepts and answering questions that do not yet have satisfactory answers.
Research work runs through Launch80, our private R&D environment. Some projects become products, some internal tools, some open source.
Approach
How an engagement works
Sometimes the best technical decision is not to build anything at all. We'd rather tell you that than sell you something you don't need.
- You explain the problem.
- We figure out what the actual problem is.
- We determine whether something should be built, bought, integrated, researched, or left alone.
- We build, configure, secure, and deploy it.
- We operate it if you want us to.
FAQ
Common questions
Something not covered here? The assistant answers from the same material and runs entirely in your browser.
Ask the assistantWhat does North Star Software actually do?
A lot, deliberately. AI, software engineering, infrastructure, cybersecurity, web and SaaS, and research and development. We are not particularly interested in putting ourselves into one neat category. If the problem involves computers, networks, software, data, AI, infrastructure, or figuring out how something works, there is a reasonable chance we can help. We build difficult things.
Can you deploy AI that runs on our own hardware?
Yes. Private deployments and local inference are a core part of what we do: models running on infrastructure you control, so your data never leaves it. That matters when the data is regulated, confidential, or simply too valuable to hand to a third party's API. We handle the hardware sizing, the serving stack, the integration work, and the operations afterwards.
Can you take over an existing codebase?
Yes, that is normal work. Inherited systems, undocumented systems, and systems whose original author left years ago. We assess before proposing anything: what it does, what it costs to run, what is actually broken versus merely ugly. Then we tell you whether it is worth maintaining, refactoring, or replacing.
Can you work with regulated or sensitive data?
Yes. Regulated and confidential data is a large part of why private deployments and local inference exist in our work at all — if the data cannot leave your control, the architecture has to reflect that from the start rather than being bolted on. Tell us what regime you are under and what the data actually is, and we will design around it.
Will we own the code and be able to leave?
Yes. You own the accounts, the infrastructure, and the code. We document what we build because documentation is not optional, and the point of documenting it is that somebody else can pick it up. If you want to take the work in-house or move to another vendor, that should be a decision rather than an escape.
What does something cost?
There is no price list, but here are rough ranges so you can work out whether it is worth a conversation. A website runs roughly $3k to $15k. A web app or SaaS build, $25k to $150k and up. An AI integration or private deployment, $15k to $100k and up. Infrastructure and operations, $500 to $5k a month. An ongoing engineering retainer, $8k to $25k a month. Research engagements are scoped against the question. What moves the number is scope, integrations, data sensitivity, and timeline. Tell us what you are trying to accomplish and we will give you a real number: hello@northstarsoftware.net
Articles & Research
Recent writing
- A Better INT4 Qwopus We Decided Not to Ship
An improved INT4 build of a 27B coder model got closer to full precision on every quality measure — and still did not ship, because it was slower and broke the fast drafter most people use.
- Baking a 27B Coder Model Down to INT4 on One Arc Pro B65
Quantizing a 55 GB coder model to fit one 32 GB Intel card, with quality bars written down before measuring — and the tool-calling regression those bars caught before release.
- We Shaped the Report Template Around One Benchmark
A data-only publishing pipeline means the template decides what a result can say. Why Launch80 tuned theirs for BetterBench, and what that costs every other kind of result.
Tell us what you're trying to accomplish.
We'll figure out the technology. Replies within 24 hours on weekdays.