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.
- qualification level
- 4
- months of programme
- 9
- taught hours
- 600 h
- credits
- 60
- competency blocks
- 5
Skilled employee and operative
Led by an instructor.
1 credit = 25 h of work, 10 of them taught.
Framework of the « Data and Annotation Operator » qualification, code TP37.
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 %)
- 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.
| Programme | Duration | Credits | Taught hours | Fees |
|---|---|---|---|---|
| Data Science programme — Bac+5 level · Master's degree | 2 years | 120 | 1200 h | 7 200 000 GNF per year |
| Advanced Excel · Short course | 2 months | 4 | 40 h | 900 000 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.
Explore next
- Administrative officerSME administrative officer: quotes and invoicing, purchasing, cash monitoring, personnel files, administrative organisation. Level 4 qualification in 9 months and 600 taught hours.
- E-commerce officerE-commerce and back-office assistant: product pages, online stock, order processing, mobile payment, customer service. Level 4 qualification, 9 months, 600 hours.
- Legal assistantLegal assistant: file preparation, deadline tracking, routine deeds, company formalities, documentary watch in the OHADA area. Level 4 qualification, 9 months, 600 hours.
- HR assistantHuman resources assistant: recruitment, contracts, personnel files, absence and leave, payroll preparation, employee relations. Level 4 qualification in 9 months, 600 taught hours.
- Graphic designerGraphic designer and production operator: layout, typography, colour management, preparing files for print and digital. Level 4 qualification, 9 months, 600 hours.
- Works team leaderWorks team leader: reading drawings, organising the work face, leading a team, safety and quality of execution on site. Level 4 qualification in 9 months and 600 taught hours.