Skip to content
AtheronLABS

You're visiting from the United States. Prices are shown in US dollars. Not right?

labs@atheron:~/industries/education/ai$ train agent --grounded

AI course tutors

Tutors that teach from your course, on a private model.

A good tutor teaches; a general chatbot hands over the answer. We build tutors that work from your course materials, follow rules your instructors set, and run on open-source models in your own cloud account, so student conversations never leave the institution.

// what matters here

What is different in this industry.

01

Teaches rather than tells

The tutor asks guiding questions and points to the reading, and instructors decide where it should stop helping.

02

Grounded in your course

Answers come from your readings, slides and problem sets, so they match how the course teaches the topic.

03

Student conversations stay with you

An open-source model in your own cloud account in Canada means conversations are not sent to an outside AI provider.

04

Instructors see the gaps

Reports show which topics students ask about most, so a lecture can address them.

// example projects

Priced examples for this service.

Each one is hypothetical, labelled as such, and priced live from our rate card. Open any of them in the estimator and make it yours.

Example projectWeb appAIOpen-source model

Course tutor on a private model

Hypothetical. Not a client, not a result.

// The problem

A college's first-year students get stuck on coursework at night, and general chatbots hand them answers that skip the learning and ignore the course.

// What we would build

A tutor that works from the course's own materials, asks guiding questions rather than giving answers away, and shows instructors where students struggle. It runs on an open-source model in the college's own cloud account in Canada, so student conversations stay with the institution.

// What is in it

  • Grounded in the course's own readings, slides and problem sets
  • Guides students step by step instead of handing over answers
  • Instructors set the rules per course and per assessment
  • Reports on the topics students ask about most
  • An evaluation set built with instructors, run before every change

// Stack

  • Llama or Qwen (open-source)
  • vLLM
  • PostgreSQL with pgvector
  • Learning platform interface (LTI)
  • AWS (Canada region)

// estimate

Build
≈ US$111,000 to US$169,000, delivered within 22 weeksCAD 157,700 to 241,100
Hosting
≈ US$6,710 a monthCAD 9,550 a month
Support
≈ US$2,500 a monthCAD 3,565 a month

Prices in your currency are estimates from today's Bank of Canada rate. All invoicing is in CAD or USD.

AI route
Open-source model
Model running cost
≈ US$6,530 a monthCAD 9,300 a month

Timeline by milestone

Discovery
2.9 to 3.4 weeks
Specification and evaluation plan
0.4 weeks
Design approved
2.4 to 4.4 weeks
Core features
1.4 to 1.9 weeks
Full build
0.9 to 1.4 weeks
Working pilot
3.9 to 6.9 weeks
Testing and fixes
0.9 to 1.9 weeks
Launch
0.9 to 1.4 weeks
Production
0.9 to 1.4 weeks

Outside our hands, and added to the calendar

LMS test access from your IT team
1 to 4 weeks

How it is paid

Deposit 20%
CAD 31,480 to 48,160
Discovery 1.8%
CAD 2,833.20 to 4,334.40
Specification and evaluation plan 10.5%
CAD 16,527 to 25,284
Design approved 1.8%
CAD 2,833.20 to 4,334.40
Core features 5.6%
CAD 8,814.40 to 13,484.80
Full build 3.7%
CAD 5,823.80 to 8,909.60
Working pilot 31.7%
CAD 49,895.80 to 76,333.60
Testing and fixes 1.8%
CAD 2,833.20 to 4,334.40
Launch 1.8%
CAD 2,833.20 to 4,334.40
Production 11.3%
CAD 17,786.20 to 27,210.40
Holdback, 30 days after launch (10%)
CAD 15,740 to 24,080
GPU time for training and testing, at cost
CAD 300. Billed up front, at cost, outside the milestones

// questions

Questions we get about this.

Can instructors switch the tutor off during an exam?

Yes. Instructors set rules per course and per assessment, including when the tutor is unavailable.

Why an open-source model for a course tutor?

Student conversations are personal information, and many institutions prefer them to stay in their own account. A frontier model is cheaper to start with where your privacy review allows it; we price both.

Does the tutor work inside our learning platform?

Yes, through LTI or your platform's own interface, so students open it from the course page.

// next

Not quite your project?

Tell us what you have in mind. We will come back to you with a range and the questions that would narrow it. Or book a call and talk it through.

AI course tutors on private open-source models | Atheron Network Labs