- BRLR$
- USD$
- EUR€
- GBP£
- MXN$
- ARS$
- CLP$
EN
During the test, candidates use an AI assistant just like they would on the job. Coodesh grades the work and shows what is behind it: what they asked for, how much came from the AI and whether they checked the result.
You build, invite and get the analysis ready to read. Nobody on your team grades tests by hand.
Start from a catalog template or describe the role and let AI build the test. You pick the format, level, skills and AI policy: which credential, how many messages, and whether the assistant can only explain or also write code.
Send a public link, an email or invite straight from your ATS. Each person takes the test in the browser, whenever it suits them.
AI grades every criterion in the brief and analyzes the conversation with the assistant. You get the score, result authorship, watch-out signals and ready-made interview questions.
No banning the tool they will use every day. Candidates use AI in the test and show how they think.
By email or link, with the time limit, the formats and what they are allowed to use. No surprises when they start.
The assistant lives inside the test: they ask, review, fix and submit, just like at work. No personal account on any AI tool needed.
After submitting, they follow the assessment status and their result in the candidate dashboard.
Every assessment comes with an analysis of the conversation with the assistant and of the final result.
Example with masked data from a real back-end assessment. The AI usage analysis does not change the score.
AI is already part of the job in every team. Find out who uses it well, with everyday tasks for each role.
Turn customer interview notes into a PRD, checking what the AI summarized.
Build an outreach sequence for a customer profile and adjust the tone the AI suggested.
Triage a ticket queue with an LLM and reply to the most critical customer.
Analyze an expense spreadsheet, explain the variances and check the numbers the AI calculated.
Reconcile records and documents, automating the repetitive part with prompts.
Write a job description and an interview guide, reviewing the AI draft for bias.
Pick the format closest to the day-to-day of the role. In all of them, the AI conversation is part of the analysis.
A browser-based coding environment with an AI coding assistant or prompt-based help.
Draw flows, architectures and processes with an AI assistant to discuss ideas with.
Written answers, analyses and documents, with the AI assistant available while answering.
The person records a presentation and can ask the AI assistant to help prepare what to say.
Get started fast with ready-made tests by role, team and level.
Describe the role and AI drafts questions, criteria and scenarios. You review and adjust.
A score per criterion with the reasoning behind each one, plus manual review whenever you want.
Prompts, result authorship, usage criteria, signals and timeline.
Learn more An AI fluency level for each person, to compare candidates and map your team.
Periodic photos, tab switching, full screen, copy and paste, DevTools, monitors and IP.
Learn more Confirm that the person who took the test is the person you hire, with deepfake detection.
Learn more Try it yourself or book a demo to see an AI-assisted test built for your role.
Start for freeSend tests and get results without leaving your ATS.
Create tests, invite candidates and check results by chatting with your AI assistant.
It is a test where candidates can use an AI assistant, just like they would at work. Coodesh grades the work and analyzes how the person used AI: what they asked for, how much of the result came from the assistant and whether they checked what they got.
Not when AI is part of the job. You set the AI policy for each question, and what is still off limits, like pasting answers from elsewhere or switching tabs, stays monitored by proctoring.
The assistant lives inside the test and runs on the Coodesh credential, so candidates do not need a personal account on any tool. You choose whether it can only explain or also write code, and how many messages they can send.
No. The score comes from grading each criterion in the brief. The AI usage analysis sits alongside it, with criteria, signals and interview questions, so you can decide with more context.
Yes. Free text, whiteboard and video formats cover teams like Product, Go To Market, Operations and CS, Finance, Back Office and People.
Yes. Start from a catalog template or describe the role and let AI draft the questions and criteria. Then review and adjust.
With proctoring (periodic photos, tab switching, full screen, copy and paste, DevTools, multiple monitors and IP) and identity verification with deepfake detection.
Yes. There are integrations with ATS such as Lever, Greenhouse, Gupy, Inhire and Recrut.ai, plus MCP to use Coodesh right inside Claude or OpenAI.
Yes. After submitting, they follow the status and result of the assessment in the candidate dashboard.
Start for free or book a demo to see an AI-assisted test built for your role.
AI recruiting platform to assess, interview and hire based on proven skills.
Create your account and send your first assessment today.
