Our mission is to make AI useful in the languages where being almost right is worthless. We start with C and C++, because that is where the cost of a wrong answer is highest and where every existing tool is weakest. We like people who are stubborn about correctness and fast about everything else.
Own the part of the agent that decides whether a C or C++ patch is actually correct.
ByteAsk builds a coding agent that has to prove its work. It runs the code under a debugger, puts it through the sanitizers, profiles it, and keeps reworking the diff until the toolchain agrees the patch is right.
You will own the part of the system that decides whether a patch is actually correct. That means running the real tools a C++ engineer runs, reading what they return, and deciding what to do next. It is an open research problem dressed as a systems problem, and today nobody has solved it: agents resolve roughly 63% of real-world issues in Python and under 10% in C and C++. As one of the first engineers you will have far more scope than the same title anywhere else.
What you’ll do
Cross-translation-unit reasoning, where every public agent scores close to zero
Turning debugger, sanitizer and profiler output into a next move
Tool selection, including knowing when the honest answer is "we don't know"
Making the loop fast enough to live inside an edit cycle
On-prem deployments
Performance optimizations
You may be a fit if
You have spent real years in C or C++ and know why the hard bugs are hard
You have read sanitizer output at 2am and gone to the codegen to settle an argument
You are comfortable owning an ambiguous problem with little guidance
You are sceptical of AI writing C++, and want to find out exactly where it breaks
You do not need an AI background. We would rather teach that half
Own the agent loop and the evals that tell us whether a change made it better.
ByteAsk is only as good as the decisions the agent makes between tool calls. You will own that loop: how it plans, which tool it reaches for, how it reads what comes back, and how we know whether a change made it better or worse.
Most of the work is building evaluations from real C and C++ bugs and then moving the numbers on them. As one of the first engineers you will own the model side of the product, not a slice of it.
What you’ll do
Building and running evals from real C and C++ issues
Improving how the agent plans, picks tools, and reads their output
Context management across large codebases and many translation units
Model selection and routing, and the cost versus quality tradeoffs behind them
Fine-tuning and data work, where it beats prompting
You may be a fit if
You have shipped LLM or ML systems that real users depend on
You treat evals as part of the product, not an afterthought
You can read C or C++ well enough to judge a fix, or want to learn fast
You are comfortable owning an ambiguous problem with little guidance
You would rather settle an argument with a measurement than an opinion
Own everything a developer touches around the agent: the CLI, the web app, accounts, and billing.
Developers meet ByteAsk through an install command, a terminal, and a web dashboard. You will own those surfaces end to end, from the first install to the day a team upgrades its plan.
The bar is a tool engineers trust on day one: fast, clear, and never in the way. As one of the first engineers you will make product decisions, not just implement them.
What you’ll do
The terminal experience: install, onboarding, and day-to-day CLI ergonomics
The web app: accounts, teams, usage, and billing
The APIs and infrastructure behind them
Instrumentation that shows where developers get stuck
Shipping small improvements every day
You may be a fit if
You have shipped full-stack products that developers use
You care about the details of a CLI as much as a web UI
You are at home across frontend, backend, and infrastructure
You write code other engineers enjoy reading
You are comfortable owning an ambiguous problem with little guidance
Put ByteAsk in front of the C and C++ engineers who need it, and bring what they say back to the product.
C and C++ engineers are sceptical of AI coding tools, usually for good reasons. Your job is to earn their trust: show real results on real codebases, show up where they already are, and turn early users into advocates.
You will sit between the community and the founders, owning content, community, and the first conversations with teams evaluating ByteAsk. What you bring back will shape what we build next.
What you’ll do
Technical content: demos, write-ups, and talks built on real C and C++ bugs
Growing and running the ByteAsk community
Early conversations with teams evaluating ByteAsk
Turning user feedback into clear product input
Measuring what actually brings developers in
You may be a fit if
You can write and talk about systems programming credibly
You have grown a developer community or audience before
You are comfortable on a call with an engineering lead evaluating a tool
You enjoy writing as much as building demos
You are comfortable owning an ambiguous problem with little guidance