Nova · your AI learning mentor

A learning pathyou can verify.

For anyone working towards a role in tech — frontend, data, cloud, ML, security, 15 roles in all — who wants to know what to learn next, in what order, and why.

Loading Nova

Ten thousand courses. No order.

Every site hands you a pile of courses. None of them starts from what you already know and tells you the shortest way to the role you actually want.

From your goal to your first course, in six steps.

OVERVIEWINTRO

Six steps

How one prompt becomes an ordered plan

A DETERMINISTIC ENGINE PICKS THE ORDER.
THE AI ONLY EXPLAINS IT.
6
STEP / 0101 / 06

Say your goal

The role you want, what you already know, and your hours a week.

1GOALCloud Engineer

2LEVELPython · comfortable

3TIME6 h / week

4STYLEvideos · projects

STEP / 0202 / 06

Find what's missing

What the role needs, minus what you have, plus what those need first.

STEP / 0303 / 06

Score every course

Five checks: gap closed, level fit, your preferences, quality, match to your goal.

Coverage0.04Level fit1.00

Preference fit1.00Quality0.70

Similarity0.31Learners' next0.24

Total0.43

STEP / 0404 / 06

Pick the fewest

The smallest set of courses that closes the gap inside your hours.

EACH PICK
MOST GAP PER HOUR

STOPS WHEN
THE GAP IS CLOSED

OR WHEN
YOUR HOURS RUN OUT

STEP / 0505 / 06

Put them in order

A topological sort over the prerequisites, then checked against real learners.

STEP / 0606 / 06

Change it when you push back

Done, too hard, too easy. The plan is rebuilt and you see what changed and why.

YOU SAIDtoo hard

SWAPPEDConfluent Kafka Tutorials

FORConfluent Developer Kafka Courses

Why trust the order we give you?

Skills

0

the skills we cover, ordered by hand.

Checkable

0of 193

“learn this before that” tested against real learners.

Confirmed

0of 188

the order held.

Disputed

0of 188

the order didn’t. A person read each one and decided.

WHY THIS COURSE

WHYTHIS?

LIVE PROFILE

FILLSAS YOU TALK

SKILL GRAPH

GAPMAP

WHEN A COURSE IS TOO HARD

THE PLANCHANGES

DASHBOARD

PROGRESSRADAR

IF THE MODEL IS DOWN

STILLANSWERS

You get a plan, the reasons behind each step, and a mentor who rewrites it when you push back.

Six thingsa course listnever does.

PATHWISE

Under the hood

The engine decides. Nova only explains.

The path is arithmetic over your profile, the skill map and the catalog. Nova reads the result and tells you why.

Profile
Skill graph
Catalog
Engine
Nova
Path
Feedback
Dashboard
Same inputs, same path, every time.
Evidence

We checked the map against two million real learners.

The map was checked against how real people learned, on Stack Overflow and Coursera. Every number carries its source and its limit.

Python before Python for Data Analysis

Verified by 2 of 2 sources

54,067 learners, 95 % in this order · Stack Overflow

Stack Overflow
support 51,421reverse 2,646confidence 95 %n 54,067
Coursera reviews
support 491reverse 77confidence 86 %n 568

2,137,848

users

Stack Overflow

The order people first asked about each skill.

Stack Overflow question order (first question per tag), users who started after both technologies existed; asking ≠ completing

72,774

learners

Coursera reviews

The order people reviewed each course.

Coursera learners 2015–2020; sequences reconstructed from review order; pseudo-users by reviewer name

Try it yourself

Startwithagoal.