Help build detection institutions can actually trust
iOriginally is a small team building a plagiarism and AI-writing detection platform — and the engines behind it — in-house. There's no sprawling job board here: just the areas we hire ahead for, and a straight way to introduce yourself.
What you'd actually be joining
An early-stage product company with the whole problem on the table — detection quality, report design, institutional workflow — and no layer of process between you and it.
The engines are in-house
Similarity, paraphrase-aware matching, anti-masking and AI-writing analysis are built and evaluated here — not white-labelled from a vendor. Detection quality is our problem, which makes it your problem too.
The platform is the product
Original View reports, verifiable certificates, LTI 1.3 launches, self-hosted deployments — the workflow around detection is where institutions live, and we build all of it.
Small team, real ownership
You own problems, not tickets. Work ships end-to-end and lands in front of real integrity offices, instructors and students — usually much sooner than you expect.
What we optimize for
These show up in the product — in the policy note on every AI screen, in metering that never bills a failed check — and they show up in how the team works day to day.
Evidence over vibes
Scores are aids for human judgment, never verdicts. Everything we ship has to show its work — matched spans, sources, sentence-level probabilities — so an integrity office can defend a decision, or overturn one.
Honest numbers
No invented statistics in the product or the marketing, no metering tricks, and comparison pages that admit where competitors lead. If we can't back a number, we don't print it.
Institutions own their data
Student work never trains models. Retention, export and deletion stay under the institution's control, and fully self-hosted deployment is a first-class option — not a concession we make reluctantly.
Ship end-to-end
One team owns the path from detection research to the PDF an integrity office files. You see your work in use — and you hear about it directly when it falls short.
Areas we hire ahead for
These aren't vacancies with closing dates. They're the areas where the roadmap will need people — and where we read every introduction that comes in.
Detection research
Similarity matching, paraphrase detection and AI-writing analysis — building corpora, evaluating detectors honestly, and hardening them against homoglyph and masking tricks.
Full-stack engineering
The platform around the engines — exact-layout report viewers, submission pipelines, LTI 1.3 integrations and self-hosted deployments — across the web app, APIs and document pipeline.
Institutional success
Onboarding, training and policy guidance for integrity offices — helping institutions use similarity and AI signals responsibly, from first evaluation through exam season.
How introducing yourself works
Write to us
Email careers@ioriginally.com with whatever shows your work best — a repo, a paper, something you shipped, or a straight letter about what you'd want to build here.
We read it
Introductions go to the team, not a tracking system. We read every one, and we reply when we can see a fit — for a role now, or for one we hire ahead for.
We talk about the work
If there's a fit, the conversation is about the problems themselves — detection quality, the platform, institutions — not whiteboard riddles.
Think you should be here?
Tell us what you'd build. A short, specific introduction beats a long CV — and the team reads all of them.