Introducing Borderless Roles: AI Job Search

Why we built a job search

We build products for other people all day. This one started as a problem in our own search.

Everyone on the team has looked for a job across borders. The mechanics are always the same and always worse than they should be. Open a board. Filter by remote. Open twenty postings. Find out that "remote" means remote within the United States. Open twenty more. Find out that "visa sponsorship available" means "not for this role." Close the tab and start the next batch.

The filtering is the job. The applications are the other half. Once you find a role that fits, you get to spend an evening retyping the same information into a form that asks the same six questions in a slightly different order.

So we built the thing we wanted. It is called Borderless Roles, and it is live today. It reads every listing before you do, and it does the repetitive part of the application without taking the decision away from you.

The fine print is the problem

A job board shows every listing and leaves the eligibility check to you. That is the whole design. The board's incentive is inventory, so it wants as many postings as possible in front of you, and the two lines that decide whether you can apply sit near the bottom of the page.

Those two lines are:

1. Where the employer will actually hire. Remote is a location policy, not a permission slip. Plenty of remote roles only hire in one or two countries, and the ad rarely says so in the title.
2. Whether it sponsors a work visa. A posting can mention sponsorship in the boilerplate and still rule you out three paragraphs later. Or it can sponsor and never use the word.

Both are readable. Neither is read. So the user does the work: open, scan, discard, repeat. That is a tax on the applicant, paid in evenings, for information the listing already contained.

What Borderless Roles does

The search reads the fine print first, then shows you the roles that survive it.

• 54,000+ open jobs and 10,000+ companies hiring, pulled from 21+ platforms and curated company career boards.
• Remote eligibility filtering. A remote role appears for your country only when its hiring zone includes your country. A role that excludes it is filtered out, not shown with a caveat.
• Visa sponsorship labelling. A role abroad is labelled as sponsoring only when the posting text itself states it. It is a signal to check, not a promise, and we never label from a government register, which only tells you where to look.
• Fit scores refreshed daily. You set your pay floor, the roles you want, and your deal-breakers once. Every listing is checked against that profile, and the roles that fit get a score that explains why.
• Real listings. The same role cross-posted on several boards, or one employer posting it city by city, collapses into one result. Anything older than 30 days is archived, so you do not spend time on a ghost job.

Listings come straight from the career pages of companies we curate, which keeps spam and fake postings out of the feed. Every board is refreshed daily and read by AI before it reaches you.

How the AI reads a listing

Every listing goes through the same extraction before it lands in the feed. The model reads the posting and pulls out the facts that decide whether an application makes sense:

• Role and seniority. Each listing is tagged with a role family and a level, so you can drop the seniority you are not applying for.
• Pay. Salary ranges are parsed into a comparable number, so a pay floor means the same thing across every board.
• Hiring zone. Remote eligibility is read from the location requirements and the fine print, not the "Remote" badge.
• Time zone. We read the zone a team hires in, so you can keep to roles whose working day overlaps yours.
• Visa sponsorship. Labelled from the posting text only. If the ad does not say it, we do not claim it.

That extraction is what makes the feed honest. A board can show you 6,200 "matches" because it counts every listing it has. We show you the ones whose rules do not shut you out.

The extension: Easy-AIpply rolled in

Easy-AIpply was our first pass at the application half. A Chrome side panel that scanned a listing once and drafted many answers from shared context. It worked, and people used it, but it was a separate product with a separate profile. You found the job in one place and applied to it in another.

It is now the application layer inside Borderless Roles. The extension opens next to any posting and works from the same profile as the search:

• Scan once. One click reads the role, the requirements, and the questions on the form. Every draft below comes out of that same scan.
• Fit score first. Before you spend a credit, the extension checks the role against your pay floor, location, work authorization, and deal-breakers.
• Drafted answers. Every question on the form gets a draft written from the listing and your profile. Add a line of context to steer it, then edit anything.
• Tailored resume and cover letter. Pick the talking points that matter for this company, choose a template, and generate a document for the role instead of a generic template.
• One-click autofill. Name, email, phone, and links fill the standard fields for free. One more click fills the drafted answers.

And the boundary we will not cross: it never submits. You review every word and press submit yourself. It does not upload files or touch captchas. It reads a page only when you click Scan, and it asks for the permissions it needs to open the side panel and read the active tab on demand, not your browsing history or other tabs.

What we kept from Easy-AIpply

The architecture survived the merge, because it was built for exactly this.

• Plasmo + React side panel. The eyes. It opens beside the form on whichever site you are on and stays open across multi-step applications.
• Go + Gemini agent. The brain. Chosen for cost efficiency and speed of iteration, and still the cheapest way we know to run this kind of drafting at volume.
• Preference memory. When you edit an answer or a document, we store field-answer preferences, preference text, and per-question instructions, then inject them at prompt time on later scans. No fine-tuning. The drafts just drift toward your voice the more you use them.
• Shared-context scanning. One scan produces the role, the questions, the resume, and the letter. Nothing gets pasted into a chat window, and nothing loses context between questions.

The change is what sits behind the extension: a search product with a real profile, real filters, and a feed that already knows where you can work.

How it works

The search is a Next.js app over curated company career boards. Each board is fetched and refreshed daily, and every listing is read by AI at ingest for role, seniority, pay, hiring zone, time zone, and sponsorship. Duplicates collapse across boards. Stale listings drop out.

The extension runs on LinkedIn, Greenhouse, Lever, Workday, Ashby, and more. It reads the active tab when you click Scan, sends the listing and your profile to the agent, and streams back the fit score, the drafts, and the documents. Standard-field autofill runs locally from your saved profile.

The two halves share one account. Set your preferences once and both the feed and the extension use them.

Search is free, pay when AI does the work

Searching and filtering are always free, and so is autofill for the standard fields on a form. Credits cover the AI that costs us to run: fit scores, written answers, tailored resumes, and cover letters. A resume or cover letter is one credit and includes ten rewrites for that job.

• Free. Unlimited search and filters, standard-field autofill, and a 7-day AI trial with 10 fit scores a day and one AI-written application a day.
• Basic, $9.99/mo. 400 AI credits a month, about 133 AI-written applications, and 100 fit scores a day.
• Pro, $19.99/mo. 1,200 AI credits a month, about 400 AI-written applications, and 200 fit scores a day.
• 90-day pass. The same product, paid once, for a focused search.

Every plan is the same product, and prices are lower in many countries. No card is needed to start.

Try Borderless Roles

Search is free. Pick your country and your feed starts with the roles that can actually hire you.

Start your AI job search for free and see what your feed looks like. If you were an Easy-AIpply user, your workflow is still here, now with a search that already knows where you can work.

Read the Borderless Roles case study for the engineering detail, or the Easy-AIpply case study for the original extension build.