Skip to main content
Institute of Advanced Technological and Commercial Studies

Careers

Data and annotation operator

No artificial intelligence model works without data prepared by people. The data operator collects, cleans, annotates and checks: a new occupation, demanding in rigour, and one of the most direct doors into the digital sector.

What working on data involves

Work starts from an annotation guideline: what you mark, how, and above all in which edge cases. Then comes production, long and steady, where consistency matters more than speed. Quality control takes an equal share: double annotation, inter-annotator agreement, review of disputed cases. On top comes cleaning — duplicates, outliers, mismatched formats — which decides the quality of everything built afterwards.

  • Apply an annotation guideline and document edge cases
  • Annotate text, image or audio with a dedicated tool
  • Clean a dataset: duplicates, formats, missing values
  • Check a batch's quality and measure inter-annotator agreement

An occupation taking root in the region

African language projects

Building corpora in Soussou, Pular or Wolof requires annotators who speak those languages. Linguistic skill becomes a technical skill here.

Corporate data teams

Banks, telecoms, retail: before any dashboard, someone has to make reference data and history reliable.

Specialist service providers

Regional providers supply annotation to research teams: an open market that hires on demonstrated rigour.

The programme data

Four values, one source. They feed the qualification page, the application form, the equivalence file and the final certificate without ever being re-entered. That single source is what makes two catalogue programmes comparable: the figures carry the same meaning and the same unit everywhere.

Skilled employee and operative

qualification level
4
months of programme
9

Led by an instructor.

taught hours
600 h

1 credit = 25 h of work, 10 of them taught.

credits
60

Framework of the « Data and Annotation Operator » qualification, code TP37.

competency blocks
5

What each of the blocks covers

For each block the framework states the ground covered and the taught hours devoted to it. That precision serves at the orientation stage, then during the programme, then on exit when you have to describe what you can do. One vocabulary throughout, that of the occupation as companies practise it.

Data and protocols
150 (25 %)
Entry, annotation and tooling
170 (28 %)
Quality control and consistency
150 (25 %)
Confidentiality and data ethics
90 (15 %)
Professional conduct and employability
40 (7 %)
Total: 600 taught hours
  • Data and protocols · 150 h — Data types, formats, annotation protocol, guidelines, edge cases
  • Entry, annotation and tooling · 170 h — Annotation tools, shortcuts, throughput, text, image, audio
  • Quality control and consistency · 150 h — Inter-annotator agreement, sampling, spotting inconsistency, escalation
  • Confidentiality and data ethics · 90 h — Personal data, secrecy, annotation bias, refusing questionable work
  • Professional conduct and employability · 40 h — Rigour, consistency, remote working, job search

Entering the data field

This table sets side by side what genuinely prepares for this role. Credits and taught hours measure the gap between two options; fees relate that gap to a budget. Those are the three questions a candidate asks, and the three answers sit on a single row.

ProgrammeDurationCreditsTaught hoursFees
Data Science programme — Bac+5 level · Master's degree2 years1201200 h7 200 000 GNF per year
Advanced Excel · Short course2 months440 h900 000 GNF in total

What comes up in an admissions meeting

Do I need a mathematics background?

The role rests on method and consistency rather than calculation. The framework installs the statistical notions useful for quality control, from scratch. Those wishing to move towards analysis find the continuation in the data and applied AI technician qualification.

Is the work done remotely?

Often yes, provided data access is properly controlled. That is what makes the occupation reachable from Kankan, Labé or N'Zérékoré as much as from Conakry, given a stable connection and a properly secured workstation.

Data and annotation operator — Careers | IHETC — IHETC