01The comparison

Where do your hard questions go today?

Chatbots hallucinate, communities answer on their own schedule, and senior consultants price by the hour.

~$75/month · billed yearly ($899)

Save 24% vs $99 month to month.

Everything included

15+ senior engineers on call
Answers you can trust, live
Live Q&A twice a week
Written answers, any day
Resume & project reviews in 48–72h
A production blueprint every month
Courses & workshops

The whole market, same seven benefits

Towards AI MentorshipFrom $75/month 7/7

ChatGPT & ClaudeFrom free 0/7

Discord & RedditFree 0/7

A single mentor$120–450/month 0/7

A senior consultant$200–500/hour 2/7

IncludedLimited or billed extraNot offered

Bought separately: $2,000+ a month. Here: about $75 a month.

What you get	ChatGPT & ClaudeFrom free	Discord & RedditFree	A single mentor$120–450/month	A senior consultant$200–500/hour	Towards AI MentorshipFrom $75/month
Senior engineers on call	✗	✗	~	✓	✓15+ engineers
Answers you can trust	✗	✗	~	✓	✓Accountable, live
Live Q&A on your project	✗	✗	~	~	✓Twice a week
Written answers, any day	~	~	✗	✗	✓Included
Resume & project reviews	✗	✗	~	✗	✓In 48–72h
Production blueprints	✗	✗	✗	~	✓One a month
Courses & workshops	✗	✗	✗	✗	✓Included

Swipe to compare →

✓ Included~ Limited or billed extra✗ Not offered

A senior AI consultant runs $200–500 an hour, with retainers from $3,000 a month. A single mentor runs $120–450 a month, with resume reviews billed $150–300 each, and workshop seats sell for $200–500. The mentorship includes all of it for $99 a month, about $75 a month on the yearly plan.

Get the senior team → $899/year See everything included ↓

02The value

Build like the top 1% of AI talent, for less than a gym membership

The top 1% have senior engineers catching their mistakes while they're still cheap to fix. From $75 a month, that team is behind you, and everything it delivers stacks.

A senior team on tap

$200–500/hour elsewhere

Consultant-grade judgment: live twice a week, in writing any day.

Assets that compound

New every month

Every month: a production blueprint, a guest workshop, expert reviews. The $199 foundational course from day one. Nothing resets.

A price that stays small

From $75/month

Bought separately: $2,000+ a month. Here: $75 a month billed yearly, or $99 monthly.

Senior AI expertise was an enterprise line item. Now it's from $75 a month.

Start compounding → $899/year Meet the engineers behind it ↓

Watch

The mentorship, explained

The Towards AI Mentorship for AI Engineers video poster

03For you

Does one of these sound like you?

Four situations, one fix. If you recognise yours, the senior team slots straight in.

The only AI person at the company

You make architecture calls nobody checks. Now a senior team does: before the sprint, not after the incident.

Building solo or freelance

Clients assume there's a team behind you. Now there is: review, patterns, a sanity-check before you sign.

Engineer moving into AI

You have the software depth, not the AI judgment. The blueprints supply the judgment; the reviews build the proof.

Courses done, career not started

You can't get experience without a job. The builds are the experience: shipped systems, expert feedback, a resume the hiring side reads.

04The offer

Everything inside the mentorship

Two jobs: ship the project in front of you, land the role after it. Everything inside maps to one of them.

8+Live callstwice a week

Any dayWritten answerswhatever your timezone

1×Guest workshopmonthly, recorded

1×Production blueprintmonthly, handed over

For the project on your desk

Any question, an engineer's answer: live or in writing

Live Q&A, twice a week

30 minutes per call

An open call with a senior engineer: an architecture decision, a bug, a career move. Ask, or just listen.

Async help, any day

Answered in writing

Post in the members' channel; an engineer answers in writing, whatever your timezone. See a real thread ↓

“Should I fine-tune, or is this a retrieval problem?”·“My agent passes tests and breaks with real users, where do I start?”·“Two offers, which stack matters more in three years?”

For your job search and your growth

Proof, feedback, and people who hire

Expert reviews

Written, in 48–72h

Resume, LinkedIn, portfolio, or a project: written feedback from people who know what the hiring side discards. Resubmit as your work evolves.

Industry workshops

Monthly · recorded

Live sessions with engineers from companies like Google and OpenAI, and with senior AI engineers in our network like Paul Iusztin.

A production blueprint

Monthly · handed over

Best practices we've learned, or a full system we've built for clients, handed over with the judgment calls behind it. First up: the exact setup our own engineers use to work with agents.

Member bonuses

The curriculum comes with the team

What	Full price	As a member
10-Hour LLM Fundamentals video courseFive in-depth 2-hour video sessions, from a basic prompt to a full production rollout	$199	Included from day one
Full Stack AI Engineering · Agent Engineering · Master AI for Work	$349–499 each	25% off, always
New courses, while we build them	Full price at launch	Alpha access at 50–90% off, in exchange for the feedback that shapes the course. Agent Engineering was built this way.

Earned pathways

Standout members earn internships with our team and publication to Towards AI's 500,000+ readers: a public credential with your name on it. Earned, never promised.

How pathways are earned

Standout means consistent shipped work: blueprint builds finished and shared, thoughtful answers in the community, write-ups that hold up to review. When we see it, we reach out. The pathways: an internship with our engineering team, articles published under your own name in the Towards AI publication, and paid course development with us. Never guaranteed, always earned.

05Your team

Meet the engineers and mentors on call

Louis-François Bouchard

Co-founder & CTO · RAG & agents

Ex-PhD at Mila · co-author of Building LLMs for Production · keynote speaker, AI Engineer World's Fair

Louie Peters

Co-founder & CEO · Industry & career strategy

Ex-J.P. Morgan VP · leads Towards AI's enterprise AI work and industry newsletter · hiring engineers and advising the companies that hire them

Samridhi Vaid

ML Engineer · Multimodal LLMs & evaluation

Built an AI scribe used by 200+ physicians across 105 facilities · agent-based eval pipelines, deployed on GCP & AWS

Omar Solano

AI Solutions Engineer · LLM applications

Builds Towards AI's production LLM systems · workshop instructor, AI Engineer World's Fair

Fabio Chiusano

AI Engineer · NLP & retrieval systems

MSc, Politecnico di Milano · previously Machine Learning Reply · writes one of the most-read NLP newsletters in the field

Rucha Bhide

Mentorship Program Manager · Portfolio & communication

Runs the program day to day · reviews documentation and project write-ups so strong work reads as strong

Plus Jaiganesan N (AI Engineer, RAG & agents) and 8 more engineers, mentors, and reviewers.

06Trust

Why AI leaders and students trust this team

The same team these leaders endorse for AI education is on call inside the mentorship.

Jerry Liu
Co-founder and CEO, LlamaIndex

Alex Volkov
AI Evangelist, Weights & Biases · ThursdAI

Sander Schulhoff
Founder and CEO, Learn Prompting

Student reviews

What students say about learning with this team

A comprehensive course!

Towards ai agentic AI engineering course was a wonderful learning experience. Joined it because I used to follow Paul Lustin and was a big fan of his writing. And the course didnt disappoint. It covered everthing about building agent apps ground up. Right from things like how to implement tool calling by hand, implementing ReAct loops, to skills that are neeeded for taking your app to production, like how to do evals and deploy to production using ci\cd. Also both Paul and Louis-Francois were super responsive and helpful during the course for any questions or issues that we face. Highly recommended!

Indraneel Rao · Agent Engineering

Complete and practical

Super comprehensive with a good balance between theory and hands-on. It also is way more affordable than many other options out there. What makes the course stand out is that it doesn't stop at the basics; it actually goes into design, deployment and optimization. A great choice to master AI engineering.

Mikhail Rybalchenko · Full Stack AI Engineering

Excellent course, highly recommended

Excellent in depth handling of tradeoffs in evaluating and deploying agent based solutions. A useful mixture of theory and practice, learnt the hard way by expert practitioners. I've used a lot of the advice in building our enterprise practices to agent use in my organisation.

Cathal Curtin · Agent Engineering

Highly recommend this course

This course is highly valuable. You can learn a lot about different critical aspects of LLMs and RAG. Before taking it, I used to wander through endless resources, most of which were, in fact, low quality. One of the things I love about the course is that it's not just a series of video lectures; it includes many code notebooks with detailed explanations, allowing you to learn by coding and engaging, which is the best approach. In order to create a good generative AI app with LLMs, a person should be knowledgeable about effective prompting, choosing the right tools, data collection and curation, vector databases, query engines, building RAG systems, creating agents, evaluation methods, and deploying efficiently. This course covers ALL of them.

Farhad Dalirani · Full Stack AI Engineering

Must Have course to learn Agentic AI Engineering

This is one of the best online course available on internet and I am really enjoying learning from this course and putting it into practice. Thanks Towards AI team for this course. Really grateful to you for the support.

Vinay Gupta · Agent Engineering

Production-Grade Agentic AI Starts Here

This course goes far beyond theory, providing deep, practical experience in building agentic AI systems that actually hold up in production. It’s an excellent bridge from experimental LLM projects to real-world AI engineering, covering everything from architecture patterns to evaluation and reliability.

Sean Myers · Agent Engineering

07Inside the channel

What #ask-an-engineer looks like

Our engineers answer in writing, any day of the week. A member's RAG evals score high while real users complain; Fabio catches why.

#ask-an-engineer · members' channelAnswered by Towards AI engineers

MemberTue 9:14 AM

Stuck on something weird. Our RAG assistant scores 0.91 faithfulness and 0.87 answer relevancy on our eval set (120 questions). But real users keep reporting wrong or “not found” answers and when I rerun their exact queries, retrieval pulls the wrong chunks. I've re-embedded with a newer model and added a reranker. Barely moved. Chunking is 512 tokens, 50 overlap. What am I missing?

Fabio ChiusanoAI Engineer, Towards AITue 11:02 AM

Before you touch the pipeline again, how was the eval set built? If those 120 questions were generated by an LLM reading your chunks and writing a question per chunk, your evals are circular. Every question is guaranteed to have exactly one chunk that answers it, phrased in the document's own vocabulary. You're not measuring retrieval, you're measuring “can we find a chunk when the question was written from that chunk.” That's always ~0.9.

Real users do the opposite: they ask in their vocabulary, about problems rather than documents, and the answer often spans two chunks or isn't in your corpus at all. Do these three things in this order:

Pull 50 real queries from your production logs and label them by hand. That's your eval set now. Expect scores to drop hard, that drop is your actual gap, finally visible.
Read the failures for vocabulary mismatch (users say “refund,” your docs say “reimbursement”). If that's the pattern, a query-rewrite step or hybrid search (BM25 + dense) will move you further than any embedding upgrade.
Count how many failing queries have no answer in the corpus at all. If it's over ~20%, your problem is content coverage plus an honest “I don't know” path, not retrieval.

Don't change your chunking. 512/50 is fine, and it isn't the variable that changed between eval and production. The eval set is.

MemberTue 11:26 AM

…the eval questions were LLM-generated from the chunks. Just ran 40 real queries from logs: faithfulness dropped to 0.62. Vocabulary mismatch everywhere. Two days I've spent swapping embedding models.

Fabio ChiusanoAI Engineer, Towards AITue 11:31 AM

There it is. Rebuild the eval set from logs first, then change one thing at a time against it. This is the most common trap in production RAG, you're in good company.

Every question gets an engineer's answer.

The next one could be yours.

Start asking → $899/year

08Pricing

One price. Everything.

Every member gets the full mentorship. The only choice is how you pay.

Yearly Save 24% · $289 off

$1,188$899/year

About $75 a month, billed once a year.

Twelve months for the price of nine. 30-day money-back guarantee.

Start yearly → $899/year

Monthly

$99/month

$1,188 over a year.

The full mentorship, month to month. Cancel anytime.

Start monthly → $99/month

09Questions

Frequently asked questions

What if I don't have questions every month?

The core arrives whether you ask or not: eight live Q&A calls, a new production blueprint, and a workshop every month. The async channel is there for the day the wall appears.

How is this different from a $150/month mentor?

A mentor is one person with one specialty. This is a team of 15+ engineers and mentors: live Q&A twice a week, a monthly production blueprint, workshops, and expert reviews. One subscription, a different specialist for each kind of question.

What if I can't make the live calls?

Two run every week, the async channel answers any day in any timezone, and workshops are recorded.

Can I cancel anytime?

Yes. The monthly mentorship can be cancelled at any time. The yearly plan carries a 30-day money-back guarantee.

How does yearly billing work?

You pay $899 once a year, which works out to about $75 a month, 24% less than paying month to month. It comes with a 30-day money-back guarantee.

When do live sessions run?

Twice a week at rotating times, so a call lands in your waking hours wherever you are. Miss one and the next is days away, with the async channel in between.

What happens to my course access if I cancel?

Course access is active while your mentorship is active. If you cancel, you can upgrade the courses you've been taking to lifetime access at a reduced price.

Where does the blueprint code live?

Handed over as working material: code, prompts, configs, and the practices around them. What you build on top is yours, in your own GitHub.

How do the internship and publication pathways work?

Earned by strong work in the monthly builds and community: internships with our team, publishing in the Towards AI publication, course development. Never guaranteed.

How is this different from buying a course?

Courses teach; the mentorship keeps you building with a team behind you. The $199 foundational course is included, but the core is live Q&A twice a week, the async channel, the monthly blueprint, and expert reviews, all compounding the longer you stay.

Still deciding?

Ask us whether the mentorship fits your situation, and we will answer honestly.

Ask us directly →

Stuck on your AI build? Every decision, verified by a senior engineer