Skills inventory

What you will be able to do

Make an honest first list of what you can do today, in the five areas that AI roles need. You use this list to choose a role, and to plan what to learn first.

Before you startWhat to know and to have readyAbout 30 minutes

What you need to know

• You completed Module 1. You know the main AI roles and their coding levels from Technical and adjacent routes: How much code does each route need?.

What you need

• Your notes.md file in the ai-career folder. You made it in AI in everyday products: Set up your course notes. If it is missing, make it again as in that section. This lesson adds the heading ## Skills inventory.

Five areas, not one

People often ask one question: “Can I code?” But the Larkfield team in One project, many roles needed much more. Grace did not code, and the project would have failed without her. AI roles use five areas of skill:

1. Programming and data tools: spreadsheets, SQL, Python, and building software.
2. Mathematics and statistics: percentages, averages, probability, and how to tell a real difference from chance.
3. Domain knowledge: deep knowledge of one field of work, such as accounting, nursing, law, retail or customer support. You learn more about it in the next lesson.
4. Communication: explaining clearly, writing, listening, and saying “I am not sure” when you are not sure.
5. Project delivery: finishing work with other people: planning, deadlines, handling problems, and checking quality.

Different roles need different mixes. An AI engineer needs a lot of area 1. An AI product manager needs a lot of areas 3, 4 and 5. Nobody needs everything at the highest level.

Three levels

Use three levels for each area. Each level describes what you have done, not what you feel.

Area	Beginner	Developing	Demonstrated
Programming and data tools	Uses spreadsheets for lists and simple sums.	Uses formulas, filters and pivot tables, or has written small scripts or SQL queries with help.	Has written code or queries that other people used, without step-by-step help.
Mathematics and statistics	Calculates percentages and averages.	Reads charts and simple statistics correctly, and knows that a small sample can mislead.	Has used statistics to make or check a real decision, and can explain the result.
Domain knowledge	Knows the field as a customer or a student.	Has worked in the field for a short time, and knows its main processes.	Has done the work for years. Knows the rules, the exceptions and why they exist. Colleagues ask them for help.
Communication	Explains things to friends and classmates.	Has written reports or presented at work or university.	Has explained difficult or technical results to people who make decisions, and they acted on them.
Project delivery	Has finished individual assignments.	Has worked in a team on a project with a deadline.	Has led or co-ordinated a project from start to end, including problems on the way.

A level is “demonstrated” only if you can point to the thing that you did. In Evidence over self-ratings, you test your levels with small tasks.

Worked example: Daniela, accountant

Daniela is fictional. She has worked as an accountant for seven years. She closes the monthly accounts for a company with 300 employees. She thinks: “I have no AI skills, so I start from zero.” Here is her inventory:

Area	Level	What she has done
Programming and data tools	Developing	Builds complex spreadsheets with lookups and pivot tables every month. Has never written code.
Mathematics and statistics	Developing	Checks accounts for unusual numbers every month. Has not studied statistics since university.
Domain knowledge	Demonstrated	Seven years of accounts, invoices, tax rules and audits. Trains new team members.
Communication	Demonstrated	Explains the monthly results to the finance director, and answers the auditors’ questions.
Project delivery	Demonstrated	Led the move to a new accounting system: three months, five people.

Daniela does not start from zero. She has three areas at “demonstrated”, and two at “developing”. Her real gap is programming, and some statistics. That changes her plan: she does not need to learn “everything about AI”. She needs a clear path from “developing” to “demonstrated” in two areas.

Look at the domain row. A new graduate in computer science with excellent Python is a beginner in that row. That person cannot know in a few weeks what Daniela knows about invoices and audits. The next lesson is about why this matters so much.

Guided practice: your first inventory

1. In notes.md, add a heading: ## Skills inventory.
2. Copy the template below under it.
3. For each area, write your level: beginner, developing or demonstrated.
4. For each area, write one thing that you have done that shows the level.
5. If you cannot think of a thing that you have done, lower the level by one.

| Area                        | Level | What I have done |
|-----------------------------|-------|------------------|
| Programming and data tools  |       |                  |
| Mathematics and statistics  |       |                  |
| Domain knowledge (field: …) |       |                  |
| Communication               |       |                  |
| Project delivery            |       |                  |

Check: every row has a level and one concrete thing. “I am good at maths” is not a concrete thing. “I calculated the monthly sales targets for my team of 8” is.

Independent variation: fill in a second inventory for a person that you know well, for example a colleague or a family member, from another field. Do they start from zero?

Common mistake

Rating yourself by confidence instead of by work

Symptom: your inventory has “demonstrated” in programming because you finished an online video course. Or it has “beginner” in communication although you present to customers every week.

Cause: people rate how confident they feel. Confidence is not a good measure. Many beginners rate themselves too high in new technical areas. Many experienced people rate themselves too low where their skill feels normal to them.

Fix: for each row, ask: “What did I do, and who used the result?” Rate the result, not the feeling. A finished course is learning. A thing that you built and somebody used is evidence.

Check your understanding

A nurse with 12 years of experience writes 'beginner' in all five areas, because she cannot code. Which rows are probably wrong, and why?

Show the answer

Probably domain knowledge, communication and project delivery. Twelve years of nursing is deep domain knowledge: clinical processes, safety rules, how patients and doctors talk. Nurses explain difficult things to patients and families every day, and co-ordinate care with many people under time pressure. Only the programming row is probably “beginner”, and maybe the statistics row.

Check your understanding

Why does the inventory have five areas instead of only 'technical' and 'not technical'?

Show the answer

Because AI roles need different mixes, and the five areas show which mix you have. A person with strong domain knowledge and communication can fit a product or evaluation role soon, even with little programming. With only two categories, that person would see only “not technical”, and conclude that they have nothing to offer.

Recap

• AI roles use five areas: programming and data tools, mathematics and statistics, domain knowledge, communication, and project delivery.
• Rate each area as beginner, developing or demonstrated, by what you have done, not by how you feel.
• An experienced professional with no code often has three areas at “demonstrated” already.

Next, you see why domain knowledge is a head start, and how people from other fields use it in AI.

Next: Your experience is a head start