Skills

MLOps: putting a model into production and keeping it there

A model trained on a laptop is a demo; a model served, versioned and monitored is an asset. MLOps covers that transition: containerisation, service exposure, drift monitoring, incident recovery and compute cost control.

What separates a demo from a service

A model in production lives in a moving world: input data drifts, volume doubles on a Monday morning, a dependency changes version. MLOps organises that reality — every model carries a version number and its training set, every prediction leaves a trace, an indicator flags drift before a user complains, and rolling back to the previous version takes minutes. That is what makes artificial intelligence usable by an organisation rather than by one person.

Package a model as a container and expose it as a queryable service
Version the model, its training data and its evaluation together
Monitor input drift and trigger retraining when it appears
Roll back to the previous version of a model within minutes

The modules that carry it

A cross-curricular module spread over several years reads « Years 1 to 3 »: it runs alongside the whole programme rather than a single term.

Teaching block	Programme	Taught hours	Year
Deployment and monitoring	Artificial Intelligence programme — Bac+5 level	140 h	Year 2
Compute cost and technical trade-offs	Data Science programme — Bac+5 level	100 h	Year 2
Deployment and operations	Computer Science programme — Bac+3 level	180 h	Year 3
Virtualisation and containers	Networks and Cloud	180 h	Year 2

How far you go

You run a supervised, versioned and reversible model pipeline in production, with its compute cost tracked. The ambition of a skill is measured by the act it authorises, then by the volume behind it.

teaching blocks: 4
of taught instruction: 600 h
credits in total: 60
professional qualifications concerned: 3

The programmes concerned

Fees are stated in Guinean francs: per year for a long programme, as a single amount for a short course.

Programme	Level	Duration	Fees
Artificial Intelligence programme — Bac+5 level	Master's degree	2 years	7 800 000 GNF per year
Data Science programme — Bac+5 level	Master's degree	2 years	7 200 000 GNF per year
Computer Science programme — Bac+3 level	Bachelor's degree	3 years	5 400 000 GNF per year
Networks and Cloud	Bachelor's degree	2 years	5 600 000 GNF per year

The qualifications that require it

A professional qualification is prepared for one precise occupation, with a stated volume and level.

Professional qualification	Target occupation	Level	Taught volume
AI Project Manager	AI project manager, data lead	6 · Manager and professional	800 h
Infrastructure and Cloud Manager	infrastructure manager, operations manager	6 · Manager and professional	800 h
Data and Applied AI Technician	data technician, AI solutions integrator	5 · Senior technician and supervisor	700 h

Three questions, three answers

Is MLOps for developers or for data scientists?

For both, and that is exactly its purpose: it brings together the operational discipline of one and the model knowledge of the other. The blocks come from both programmes.

Is compute cost covered?

Yes: a full hundred-hour taught block is devoted to it in data science, covering the trade-off between model size, expected latency and infrastructure bill.

Where are models hosted during the training?

On an execution environment provided by the institute, sized for the coursework and administered by its technical teams.

Explore next

See all 56 pages

Start an application