Does AI grade the candidate's report?

No. Fixed rules score the deliverable against an expert-defined answer space, and no AI model grades the response. Every prompt is kept for human review. People make the hiring decision.

Our finance team lives in Excel. Why is there no spreadsheet to build?

The task tests what happens after the numbers are pulled: checking them against the source data and writing the conclusion with an AI assistant. Spreadsheet exports open as tables beside the brief. Candidates check them and write a report.

Do the tasks include errors for candidates to catch?

Some do. Incorrect totals or unsupported conclusions test whether candidates check the source data. Scoring reflects whether they catch these errors.

Will our legal and compliance teams accept automated scoring of an AI task?

Scoring is deterministic: fixed rules against an expert-defined answer space, documented in the Technical Manual. No AI or machine learning makes automated hiring decisions. Business tasks always run with the AI Assistant on, and every prompt is saved for review. Bring your legal and compliance teams in early and we will supply the documentation.

Which roles does this cover, and can we assess our existing teams?

Role-tailored tasks cover sales, marketing, finance, operations and customer roles, and general tasks apply to any role. Our team confirms the fit of a task to your role before you send it. Skills Intelligence runs the same business tasks for employees to establish a baseline. For other functions, see Assess AI skills across your business.

We also hire engineers. Does this fit alongside our technical assessment?

Yes. Codility assesses engineering candidates on real-world tasks and in live technical interviews on the same platform, and financial services organizations run both. See how financial services organizations use Codility

See which finance and operations candidates check the numbers AI gives them

Assess finance and operations candidates using source data and Codility’s AI Assistant. See whether they check the figures and correct the errors.

Request a free trial

Take the product tour

Validated assessments since 2009 SOC 2 Type II ISO 27001

Month-end close · Business day 3Sample task

variance_commentary.docxDrafted by AI Assistant

September opex closed at $2,184k, 3.2% under budget, with travel $41k below plan.

Candidate’s edit

3.4% over budget. The export shows +$71k on a $2,113k budget.

sep_export.xlsx

cost_ctrbudget_kactual_kvar_k travel96.055.0-41.0 software310.0393.0+83.0 other_opex1,707.01,736.0+29.0 total2,113.02,184.0+71.0

Screen reportSample91%

Checking AI outputExpert

Working with source dataAdvanced

Fixed rules scored the corrected memo. All 6 prompts saved for review.

AI now prepares the first draft. Your interviews don’t test who can check it.

01

Unchecked AI figures can reach forecasts, board decks and payment runs.

02

Variance commentary written by AI reads well even when it explains the wrong number.

03

Tight deadlines, like month-end, are when an unchecked figure can do the most damage.

04

Operations teams triage requests against service levels and policy, and every call they make has to be explained.

Start with a task built on real business data

Choose general business tasks or tasks tailored to finance and operations. Candidates might check a summary against spreadsheet data and submit a corrected report, or prioritize requests against stated criteria. Tasks take 5 to 60 minutes.

Finance task brief with an attached spreadsheet open as a table beside it

Assessment report view with the candidate's submitted verified report on the left and the complete AI Assistant conversation on the right

See how the candidate checks the data

Spreadsheets open as tables beside the brief, with fixed headers. Some tasks include incorrect totals or unsupported conclusions for candidates to catch. Every assistant exchange is saved.

Review the report and the check behind it

Fixed rules score deliverables against expert-defined criteria. No AI model grades responses. Controllers and finance hiring managers review scores, assessed skills, integrity risk ratings and every prompt before deciding whom to interview.

Business assessment report showing the task score, assessed skills, an integrity risk rating and a link to the AI activity record

Skills Intelligence report for one employee's business AI assessment, showing the score and assessed skills

Assess the finance and operations teams you already have

Use the same tasks in Skills Intelligence to establish a baseline for finance and operations teams, plan upskilling and reassess against the same criteria.

Skills Intelligence for business teams

See the report the numbers support

A modeling test shows how a candidate builds a spreadsheet. Codility adds a report checked against source data with AI, fixed-rule scoring and every prompt to review, backed by almost two decades of assessment science.

Screen for business roles

Questions finance and operations leaders ask

Review a finance task with your hiring team

Bring the role you are hiring for. We will show you the task, the report and the prompts behind it.

Talk to us