Live AI bootcamps. Real projects. People ready for what’s next.
A new world of work.Learn to build it.
From Sydney to New York, Mumbai to San Francisco and Colombo to Tokyo, the next wave of AI builders is learning by doing. We teach AI through real projects, and we are developing the entrepreneurs and builders who ship what comes next.
Sydney, San Francisco, New York, Colombo, Mumbai, Tokyo, and the world.
Technologies
we teach
we teach
ClaudeOpenAIGeminiCursorGitHub CopilotAWSAzureGoogle CloudDatabricksSnowflakePythonPyTorchHugging FaceLangChainMCPn8nMakeElevenLabsSupabaseDockerKubernetesTerraformdbtVercelPerplexity
What happens next
We turn a beginnerinto someone who ships AI.
Twelve weeks, 25 people, one engineer who does this for a living. You build something real every week, and you leave with work an employer can open and run.
60programmes, from first prompt to production
25people per cohort, so every project gets reviewed
3live sessions a week on Zoom, recorded for replay
15+countries our students learn from, from Australia to Canada
Trending courses
What people are opening this month.Six live bootcamps, 12 weeks each.
The bootcamps with the most visits in the last 30 days. Each one runs live on Zoom, and you deploy six projects along the way.
Free course handbook
Forward Deployed Engineering, the hottest role in AI.
The customer-facing role that AI labs and enterprise vendors are hiring for fastest. The handbook lays out the twelve weeks, the six projects you deploy, and what the job pays in six markets.
- Week-by-week plan
- Pay in six markets
Free. A few details and the PDF is yours, also sent to your inbox.
US$175k median pay in the US, Levels.fyi 2026 Popular certifications
Get certified while you learn.Free practice for Claude, AWS and Azure.
Study each exam domain by domain, then sit timed mock exams at the real length. Nova, our certification tutor, explains what you got wrong.
What we believe
AI is new for everyone. That makes this the fairest start a career has had in decades. The people who learn to build with it now, in Colombo or in New York, will shape the work that comes next. We exist to make sure that includes you.
From learner to builder
From first click to hired,or to your own company.
Follow one student through seven steps and every product they use on the way. Press play, or pick any station on the map.
I am Your route: a live bootcamp, then the hiring board.
Step 1 of 8
01Phase 1 · KnowledgeWeek 0
You, the learner
Arrive as you are. No code and no background needed. The skill check places you at Foundations, Practitioner, Builder or Advanced.
You leave with
- A placement in the right bootcamp
- Your starting level
- A clear first week
Products this student has used2 of 13
Skill checkProgramme bookletLive bootcampAI TutorCommunityProjects and reviewsCompetitionsPortfolioCertificateCertification prepInternshipHiring boardStartup School
At this step: Skill check (Free, five minutes), Programme booklet (Free PDF)
The tools you learn on.
The stack across our 59 programmes. The number is how many programmes teach it.
Python39 programmes
Claude34 programmes
ChatGPT34 programmes
GitHub17 programmes
Docker16 programmes
Gemini11 programmes
AWS10 programmes
Vercel10 programmes
Microsoft Copilot10 programmes
PostgreSQL9 programmes
Jupyter9 programmes
TypeScript8 programmes
pandas6 programmes
Hugging Face6 programmes
Cursor6 programmes
Terraform5 programmes
PyTorch5 programmes
Kubernetes5 programmes
Supabase4 programmes
Azure4 programmes
LangChain2 programmes
FastAPI2 programmes
NumPy2 programmesn8n1 programme
What you build.
Every bootcamp ships six deployed projects. Here are eight of them, word for word from the programme pages.
AIA RAG application over a real document set, deployed with citationsGen AI BootcampBuilt with Python, Claude, ChatGPT, LangChain, Vercel
AIThe capstone: an agent fleet doing real work on a schedule, monitoredAgentic AI BootcampBuilt with Python, Claude, ChatGPT, LangChain, Docker
Machine LearningA tabular model with a leak-free validation scheme, deployed as an APIML Engineer BootcampBuilt with Python, PyTorch, scikit-learn, pandas, Jupyter
Data ScienceA live dashboard for a stakeholder, deployed, with a recorded walkthroughData Science BootcampBuilt with Python, pandas, NumPy, PostgreSQL, Jupyter
Software EngineeringThe app under test with a CI pipeline, deployed on every mergeSoftware Engineering with AI BootcampBuilt with Python, TypeScript, GitHub, GitHub Copilot, Vercel
CloudA containerised API on a managed Kubernetes cluster with autoscaling, deployedCloud Engineer with AI BootcampBuilt with AWS, Azure, Terraform, Docker, Kubernetes
Cyber SecurityA penetration test of a lab application with a report, publishedCybersecurity BootcampBuilt with Python, Docker, AWS, GitHub, ChatGPT
AI for BusinessAn outbound system with AI-written, human-approved messages, runningGTM Engineer BootcampBuilt with ChatGPT, Claude, Gemini, Microsoft Copilot
What your fee pays for
Your bootcamp, itemised.Nothing hidden, nothing extra.
A bootcamp is a real decision. Here is everything one price covers, line by line, and what it works out at.
Included in every live bootcamp13 items
IKnowledge
- Weekly live sessionsTaught on Zoom by an engineer who builds AI for a living, recorded for replay.
- A cohort of 25Small enough that every person and every project gets attention.
- The AI TutorPractise and ask questions between sessions, in your own time zone.
- The programme bookletThe full curriculum, week by week, before you start.
IIPractice
- 10 projectsSix deployed projects and four mini builds, all yours to keep.
- A review of every buildWritten, line-by-line feedback on the work you hand in.
- A graded competitionScored against a published rubric, with a public leaderboard.
- A Square1 Ai certificateShows what you built and exactly how it was judged.
IIIImpact
- A published portfolioLive links and results on your Square1 profile.
- The internship programmeReal projects with real teams, with a mentor beside you.
- A place on the hiring boardWhere employers post roles and look for people.
- A route into Startup SchoolTurn your best project into a company with us.
- Your programme communityPeople on the same path, from day one.
Total, all of the aboveA$200
Not ready for a live bootcamp?
Learn on your own with the AI Tutor.
All 23 self-paced courses, with Nova grading every lesson, quick check and project. Lesson 1 of every course is free.
A$15a month
See the AI Tutor Programmes
Built for the AI jobs that exist now.60 programmes, nine curricula.
Pick a curriculum to see exactly what you will build in its live bootcamp, how long it takes and the tools you will use.
Live bootcampPython and Git; no machine-learning background needed.
Gen AI Bootcamp
Build and ship LLM products: prompting, retrieval, agents, evals, production.
12weeks, live on Zoom
15hours a week
25people per cohort
10projects you build
Taught with
PythonClaudeChatGPTLangChainVercel
A$200founding price, until 1 December Reserve a seat Get the booklet
What you will build
- 01Week 1Mini buildPrompt LabMini buildA prompt notebook that compares three prompts on the same inputs and records the outputs side by side
- 02Weeks 1–2Deployed projectLLM foundationsDeployed projectA deployed assistant with structured output and a test suite
- 03Week 3Mini buildEmbedding ExplorerMini buildA small script that embeds a document set and returns the closest passages for a question
- 04Weeks 3–4Deployed projectRetrievalDeployed projectA RAG application over a real document set, deployed with citations
- 05Weeks 5–6Deployed projectAgentsDeployed projectAn agent that completes a multi-step task with tools, deployed with traces
- 06Week 7Mini buildGolden SetMini buildA labelled set of questions and answers for your RAG app, with a scoring script
- 07Weeks 7–8Deployed projectEvals and safetyDeployed projectAn eval harness that runs in CI for the projects so far
- 08Week 9Mini buildCost TrackerMini buildA logging layer that records tokens, latency and cost for every LLM call to a dashboard
- 09Weeks 9–10Deployed projectProductionDeployed projectThe capstone: a production LLM product with monitoring and a cost model
- 10Weeks 11–12Deployed projectEmployer brief and hiring sprintDeployed projectThe employer brief shipped, and demo day
AI: 4 live bootcamps and 7 on-demand courses See every AI programme
Your first job
We don’t stop at the certificate.We train for what employers ask for.
We read 315 AI job ads to see what hiring teams actually write down. Then we built every bootcamp around it.
39%of AI job ads ask for work you have shipped to production.
12%mention a degree of any kind. What you built matters more.
46%of the roles are remote. Your first job does not have to be in your city.
What AI job ads ask forWhere you learn it with us
- Shipped to production 39% Your 6 deployed projects
- Agents and agentic systems 39% Agentic AI Bootcamp
- Python 38% Every engineering bootcamp
- Ownership and autonomy 37% Your capstone and internship
- LLMs 34% Gen AI Bootcamp
- Evals and evaluation 23% LLM Evals and Reliability
- A portfolio or GitHub work 17% Your published portfolio
- Monitoring and observability 17% LLMOps Bootcamp
- AWS 16% Cloud Engineer with AI
- Anthropic or Claude 16% Claude Code for Engineering
- A degree of any kind 12%
- A certification 7%
Square1 Ai research, “What AI job ads ask for in 2026”: 315 AI, ML and data roles from 2,470 postings on Hacker News “Who is hiring?” (July to September), Remotive and Arbeitnow, collected 19 September 2026. Short ads under-report tools, and Australia is barely in the sample. Read the method
How we get you hired
Four steps after your bootcamp, each one aimed at what those ads ask for.
- 01Publish a portfolio they will openYour ten projects on your Square1 profile, each with a live link, a result and a paragraph for the person hiring.
- 02Get certified, with the score to back itYour certificate links to your graded record and competition rank, so an employer sees exactly what it measures.
- 03Prove it in the internship programmeReal projects with real teams. You leave with experience to talk about and a mentor’s reference.
- 04Get seen on the hiring boardEmployers post roles and browse profiles. You apply with your work, your rank and your mentor’s note.
Live · updated Monday 13 OctoberSample data
Hiring board
42open AI roles
18remote-friendly
RAGmost asked skill
- AI engineer Posted with Square1RAGPythonLLM evals2 days
- Machine learning engineerPyTorchMLOps3 days
- Applied AI developerAgentsTypeScript1 week
Hiring? Post a role
See the hiring boardStartup School · Cohort 1 starts 2 November 2026
Build an AI company, not just a project.Two programmes. No equity taken.
Founders move through gates, not lectures. Each gate asks for evidence: interviews, users, revenue. You pass it or you don’t, and you see exactly why.
Accelerator
Twelve weeks. Four gates. One number.
For a builder with a market they can reach and no customers yet.
12weeks
4gates
16seats, pods of 8
15hours a week
A$790A$1,650
Founding price for Cohort 1. No equity.
Applications close 16 October · decisions by 23 October
The gates · 12 weeks
- Weeks 1–3 Evidence 30 logged interviews and one hypothesis proved wrong.
- Weeks 4–6 Ship A live product, a security score of 70 or more, and 10 users who did not know you before.
- Weeks 7–9 Revenue and model First paid revenue or a signed letter of intent, and a financial model that ties.
- Weeks 10–12 Scale path and capital A decision on funding, and your pitch ready for Demo Day.
- 4 Feb 2027 Demo Day Present what you built and the numbers behind it, and publish your founder record.
To apply you need
- Ten logged interviews with people in your market, before we decide
- Something you built and deployed, with the code and a live link
- A named market, and why people there take your call
- Fifteen hours a week for twelve weeks
Community
You won’t learn alone.Every programme comes with its people.
You join your programme’s group on day one. From there you share what you ship, ask when you are stuck, and compete on real problems.
Rooms and groups
Two rooms, ten subject groups, one thread every Monday.
ShipShow what you built AskGet unstuck, fast UpdateShare where you are
RDeployed my policy search agent. It cites every answer and says when it does not know.
TMy retrieval misses tables in PDFs. What chunking worked for you?
MWhat are you shipping this week? Post one line and one link.
Example posts
Competitions · Season 1
Real problems. Public data. Cash prizes.
- Climate techSolar BritainForecast Britain’s solar output every half hour from open weather data.MAE
- Climate healthClean Air USAPredict daily fine-particle pollution at 250 US monitoring sites.MAE
- ESGEmissions GapEstimate each power plant’s yearly emissions from public records.RMSLE
- SportsNBA MarginCall the home team’s winning margin in NBA games.MAE
5 OctSeason opened2 NovPrize pool revealed29 NovEntries close7 DecResults
Enter a competition Teach with us
55%of what your course earns is yours. Graduates can apply to teach on Square1 Ai.
A note from our founder
Why community and research sit inside every programme.
Learning AI is easier with people beside you and evidence behind you. Our founder explains how the Square1 Ai community and our research shape the way you learn.
Nikhil De SilvaFounder, Square1 Ai Research
We study the work we teach.128 pieces, written to be checked.
Every programme is shaped by what employers ask for and how the field is changing. We publish what we find, with the sources and the limits, so you can check it.
Our own study · Hiring & Assessment
What do AI job ads actually ask for? We read 315 of them.
We counted what 315 AI job ads actually ask for. The job is now called AI engineer far more often than ML engineer, and “production” appears in more ads than Python.
315AI job ads read
2,470postings screened
39%ask for shipped work
Latest researchLatest first
- Is prompt engineering still worth learning for developers in 2026? Yes, as an engineering skill rather than a job title: prompt work now means system prompts, structured output, examples, tool descriptions and evals that check each change. Of 315 AI job ads collected on 19 September 2026, 17 (5%) named prompt engineering, against 73 (23%) for evals and 123 (39%) for agents.
- How do you build a production app on the Claude API? Put the API behind your own server, version the system prompt, use structured output and tool use, stream responses, retry the right errors, cache stable prompt content, log cost per request and run evals before every change. Of 315 AI job ads collected on 19 September 2026, 49 (16%) named Anthropic or Claude.
- What is the Model Context Protocol (MCP), and should developers learn it? MCP is an open protocol, introduced by Anthropic in 2024, that lets AI applications connect to tools and data through standard servers exposing tools, resources and prompts. What it is, when to build a server rather than call an API directly, and the security risks. Of 315 AI job ads collected on 19 September 2026, 28 (9%) named MCP.
- What can you actually build with multimodal AI in 2026? Document and form extraction, screenshot and UI understanding, image inspection, chart reading, audio transcription feeding a text model, and video summarisation. What each looks like, where multimodal models fail (small text, counting, spatial reasoning, confident wrong extraction), how to test on your own images, and the cost and privacy questions.
- How do you start using Claude Code on a real codebase? Start small and supervised: one scoped change in code you know, a plan before any edit, a CLAUDE.md with your test command and conventions, tight permissions, tests as the contract and a review of every diff. A practical first two weeks for one developer, and when to add hooks, skills and MCP. Claude Code was named in 27 of 416 software engineering postings collected on 29 September 2026.
For teams
Your team has AI.Few use it well.
We train your people live, on the work they already do and the tools you already pay for. Every session ends with something working: an agent, an automation or a rebuilt workflow your team keeps.
- 01 Live online A$149 a person, minimum 5 Learn practical AI tools on your team’s real tasks, from anywhere.
- 02 In person A$2,500 a day, up to 20 people A trainer in the room for a full day with your team.
- 03 Train + build from A$4,990 scoped to your workflows We train your people, then build the first workflows with them.
Company brain
Build your company brain in 12 weeks.
Your tools connected, every answer tested on your own questions, and your team trained to run it.
Your knowledge
Where answers land
Agents, picked from the pilot
Google DriveMapped
NotionMapped
ConfluenceMapped
Written SOPsNot written yet
GmailMapped · added later
HubSpotMapped · added later
ZendeskMapped · added later
HR · payroll · legalKept out
Slack or TeamsAnswers with sources
WhatsAppAnswers on the go
OnboardingMost asked in the pilot
Ticket repliesA person reviews first
ProposalsA person reviews first
Champions namedA sponsor and 10–20 people
Company brainNot built yet
Test set0 / 20 real · 0 / 10 trick
Real
Trick
Questions written · today’s answer times measuredTeach track · your peopleIn person kick-off day
Build track · with usSource map · baseline timed
1 · Audit · Weeks 0–1
We map every place your company keeps what it knows, and who can see what. You name a sponsor and 10–20 champions, and together we write 20 real questions and 10 trick ones. We time how long the 20 take to answer today.
Your time (est.): sponsor 2 h · champions 1 daySponsor signs the data agreement and scopeBaseline: today’s answer times
Where our students are
Students in 15 countries.Five continents and counting.
People learn with Square1 Ai from the United States to New Zealand. Every live session runs on Zoom, so where you live never decides what you can learn.
15countries
5continents
- AsiaBangladesh, India, Nepal, Pakistan, Sri Lanka, Uzbekistan
- EuropeLatvia, Portugal, United Kingdom
- AfricaBurkina Faso, Ethiopia
- North AmericaCanada, United States
- OceaniaAustralia, New Zealand
Next cohort · 2 November 2026
One cohort. 25 spots.Yours, if you start now.
Hold your seat at the founding price. Not sure it’s the right programme? The five-minute skill check will point you to one that is.