Skip to content

An AI CV maker that rewrites your CV for every job

Paste the job description and get a CV and cover letter written for it, in Arabic or English. Each one comes with an ATS score, the gaps behind it, and live job openings matched to your CV.

ATS Score

92
Skills88
Keywords94
Job title90
Experience100
Education100

Still missing: Terraform, GraphQL

Real jobs you can apply to today

As soon as your CV is ready we search what is actually posted right now, and return up to five roles ranked by how well you fit. Every one opens the real listing, so you can apply in the same sitting.

5

live jobs matched

Each one opens the real posting

  • Frontend Engineer · RiyadhStrong match
  • Product Engineer · JeddahStrong match
  • Full-stack Developer · RiyadhPartial match
  • UI Engineer · RemotePartial match
  • Engineering Lead · RiyadhStretch role

3 credits free every month, no card. That is 3 CVs, in English or Arabic.

Tailor your CV for roles at

  • Saudia
  • SAP
  • Siemens
  • Accenture
  • Ericsson
  • Nokia
  • Cisco
  • Huawei
  • Samsung
  • Google
  • Apple
  • Nvidia
  • Snap
  • Shell
Encrypted uploads
Transparent ATS scoring
Every line checked against your CV
Humanizer strips AI phrasing

How it works

Three steps, and the work is done

Tailoring a CV properly takes 30 to 45 minutes per application. Here it takes a couple of minutes, and you enter your information once rather than once per job.

  1. 1

    Add your CV

    Upload a PDF or Word file. If you do not have one yet, fill in a form instead.

  2. 2

    Paste the job posting

    The whole advert. We read what the role actually asks for, in Arabic or English.

  3. 3

    Get everything you need

    A tailored CV and cover letter, your ATS score and what it is missing, and up to five live jobs to send it to.

One CV, sent everywhere

Two postings for the same job do not screen for the same thing

Sent unchanged, a CV puts the same order in front of every employer, and the experience one of them is actually screening for ends up near the bottom of page two. Tarshih reads each posting and rebuilds the order around what that employer asked for. Nothing is added. It is the same record, read twice.

One CV, read against two postings

brought forwardmoved down

Frontend Engineer

  • React
  • TypeScript
  • Figma
  • Node.js
  • PostgreSQL
  • Docker
Match89

Full-stack Developer

  • Node.js
  • PostgreSQL
  • Docker
  • React
  • TypeScript
  • Figma
Match72

The same CV both times. Nothing added, nothing invented.

What you get

A tailored CV, the scores behind it, and the jobs to send it to

One run produces all three: the documents, the scoring that explains them, and the openings they are aimed at.

A CV rewritten for the posting, and a cover letter to match

Both are written against the job description you paste, in Arabic or English, and typeset properly in either. Every line traces back to something already in your CV.

Cover letter

English

Your posting asks for someone who has taken a dashboard from prototype to production.

That is the work I spent last year on, and the part I would want to talk about first.

Tailored CV

العربية

الخبرة العملية

مهندس واجهات أمامية

شركة تقنية · الرياض · 2022 حتى الآن

  • أعدت بناء واجهة لوحة التحكم باستخدام React، فانخفض زمن التحميل الأول إلى النصف.
  • نقلت 24 شاشة إلى TypeScript دون إيقاف الخدمة.
One generation, both scripts, each set in its own direction.

See what is missing before a recruiter does

An ATS score with the weighted factors behind it, a separate match score for the role, and the specific gaps that cost you the rest. Each gap comes with an honest way to close it, never a way to claim it.

ATS score

91

Job match

84

What is missing

  • TerraformRequired

    You have provisioned infrastructure by hand. Say which parts you automated, and with what.

  • GraphQLPreferred

    The API layer in your second role is close enough to name. Say what it served.

  • Team leadershipRequired

    You mentored two engineers last year. That belongs inside the role, not in a list of skills.

Ranked openings, each one a click from the posting

Once the CV is ready we search what is posted right now and return up to five roles, labelled by how much of what they ask for your CV already shows. Every row opens the original listing.

How each opening is labelled

  • Strong match

    Your CV already shows most of what the posting asks for.

  • Partial match

    Part of what it asks for is in your CV, and part of it is not.

  • Stretch role

    Past what your CV currently shows. Listed rather than hidden from you.

Each one opens the real posting

The rest of it

Start without a CV

No file to upload? Fill in a form instead and the same pipeline runs on what you entered.

Or upgrade the one you have

Upload a PDF or Word file. It is read as it stands, so nothing has to be retyped.

PDF and Word

The CV downloads as a PDF or a .docx file. The cover letter comes as a PDF.

Every version, kept

Each CV you generate stays in your account with its role, its scores and its files.

LinkedIn profile content

A headline, an About section and a paste-ready block for every role, written from a CV you tailored here. In English, because that is how recruiters search.

Included with Pro and Elite

Interview questions, answered

The questions that specific role is likely to open with, each one with how to answer it from your own experience.

Included with Pro and Elite

What it will not do

The rewrite is checked against your own CV before you see it

A model that writes well will also, given the chance, write something that is not true. That chance is taken away here: every rewritten line is read back against the facts extracted from your document, and a line that adds anything is sent back.

One line, checked

In your CV

Managed the reception desk at a dental clinic.

  • Allowed

    Ran front-of-house operations for a dental practice.

    A better name for the same work. Nothing new is claimed.

  • Sent back

    Ran front-of-house operations for a 12-chair dental practice.

    Twelve chairs is not in your CV, so this line never reaches your document.

A rejected line is rewritten and checked again, up to twice, before the CV is built.

Your documents stay yours

Uploads are encrypted in transit and at rest. Tarshih never trains models on your resume or shares it with third parties, and you can delete everything permanently at any time.

The score shows its working

One number, and the weighted factors behind it: skills at 40 percent, keywords at 25, job title at 15, experience at 12, education at 8. Then the gaps that cost you the rest, each with an honest way to close it rather than a way to claim it.

It has to read like you wrote it

The same rules run on every CV, cover letter and LinkedIn profile: no dash punctuation, no "moreover" or "furthermore", no inflated significance. They are part of how the text is written rather than a second pass over it afterwards.

FAQ

Questions, answered

See all questions
  • A credit is what you spend generating one tailored CV and cover letter — 1 credit, in English or Arabic. Credits also buy more Job Search, Interview Prep or LinkedIn Essential once your monthly allowance for one of those runs out. Free includes 3 credits a month, Pro includes 24, and Elite includes 100.

  • No. You can upload an existing resume to upgrade it, or build a brand new one from scratch. Either way, the output is tailored to the specific job you're applying for.

  • No. AI models can state things that sound plausible but aren't true, and on a CV that is the risk that matters, because it is your name on the document and you are the one who has to answer for it in the interview. Every fact is extracted from your real CV first, and every generated bullet is then checked back against those facts in a dedicated fact-check pass before you see it. Tarshih reframes and professionalizes what's true; it never fabricates.

  • It shouldn't be. Every generation follows an explicit set of rules against the habits that give AI writing away: dash-heavy punctuation, signposting words like moreover and furthermore, inflated phrases such as plays a vital role, padded lists of three adjectives, and the uniform sentence rhythm people notice without being able to name it. What comes out reads like a competent professional describing their own work, which is what it's built from.

  • Your documents are encrypted in transit and at rest. We never sell your data or use it to train models, and you can permanently delete your files at any time.

Your next application deserves better odds

Try Tarshih on your next role in under five minutes. Free to start, no credit card, no commitment.