What is the AI Project Cycle?

Suppose you have to make a birthday card for your mother. You look for ideas, list what you need, fetch the paper and colours, make the card, start again if you spoil it, and finally give it to her. Nobody taught you those steps. Your mind made a plan.

1A plan for an AI project

An AI project needs a plan too, and a much more careful one. The plan has a name.

AI Project Cycle

The AI Project Cycle is the step-by-step process followed to complete an AI project. It has six stages, and it is called a cycle because the last stage can send you back to the first.

The six stages, as a cycle

Tap a stage to stand on it, or press play and watch the work go round. Drag sideways to turn.

The work travels from Problem Scoping all the way to Deployment — and the ring does not end there. What you learn at the end sends you back to the beginning.

Six stages: Problem Scoping → Data Acquisition → Data Exploration → Modelling → Evaluation → Deployment.

2The six stages

#	Stage	What happens
1	Problem Scoping	Decide what problem you are solving and set the goal
2	Data Acquisition	Collect data from reliable sources
3	Data Exploration	Draw the data as graphs so patterns can be seen
4	Modelling	Build and choose the model that makes the decision
5	Evaluation	Test the model on new data it has never seen
6	Deployment	Put the finished model where people can use it

A way to remember the order

Scoping, Acquisition, Exploration, Modelling, Evaluation, Deployment. Read the first letters: SAEMED. Silly, and it works.

3Why it is a cycle

A straight list would end at Deployment. Real projects do not. When you evaluate the model you often find it is not good enough — so you go back and collect better data, or build a different model, and come round again. Even after deployment, people use the project and you learn something new from them.

The card you had to remake

Remember the card. If you spoiled it, you did not carry on gluing — you went back and started again. An AI project does exactly this, and that going-back is what makes it a cycle instead of a line.

Six, not four or five

Some older books stop at Evaluation and say the cycle has four or five stages. Your syllabus has six, and Deployment is the sixth. The 2026 board paper asked about it, so learn all six.

4From the board papers

Which stage of the AI Project Cycle involves testing the model on newly fetched data?
(A) Data Exploration (B) Modelling (C) Evaluation (D) DeploymentCBSE board exam 2026

1 mark

▶Show the answer

(C) Evaluation. Testing on data the model has never seen before is exactly what evaluation means. Deployment comes after that, when the tested model is put to work.

What do you do in the second and third stages of an AI Project Cycle?CBSE sample paper 2025–26

2 marks

▶Show the answer

The second stage is Data Acquisition: we collect the data the project needs from reliable and authentic sources. The third stage is Data Exploration: we represent that data as graphs, charts and other pictures so that the patterns hidden in it become easy to see.

5Check yourself

Quick Check

How many stages does the AI Project Cycle have?

Quick Check

Which stage comes immediately after Modelling?

Stage 1 — Problem Scoping 🎯