Change your trajectoryinto AI & ML
$ python train.py --data cifar10 --epochs 12
» loading dataset · 50,000 train / 10,000 val
» device cuda:0 · batch 128 · lr 3e-4 · adamw
epoch 01/12 loss 2.041 val_acc 34.2%
epoch 04/12 loss 1.412 val_acc 52.6%
epoch 07/12 loss 0.874 val_acc 70.1%
epoch 10/12 loss 0.529 val_acc 82.3%
epoch 12/12 loss 0.317 val_acc 88.4%
» saved checkpoints/resnet18-best.pt
» POST /predict · 200 · 41 ms
2.041 · acc 34.2%
AI & ML Launchpad, Cohort 1
Learn the real machine learning and AI workflow - real problems, real deployment - for absolutely free, from your first line of code.
- FREE
- 14 weeks
- Starts August 8, 2026
- When
- Saturdays & Sundays · 7:00 PM PST · 2 hrs
- Where
- Live on Zoom · 27 live sessions
Registration closes August 4, 2026
Tools & frameworks you will master
AI & ML Launchpad, Cohort 1
FREE · Live on Zoom · August 8 – November 7, 2026
Registration closes August 4, 2026
Who this is for
FOR YOU IF
- You want the actual ML workflow: messy data, real deployment, not just theory slides.
- You can commit to two live sessions a week for 14 weeks, including Portfolio Presentations.
- You're a complete beginner, or a working professional adding ML to your toolkit.
- You want to apply AI to your own field — medicine, engineering, whatever it is — not a one-size-fits-all project.
- You want a portfolio project you can show in an interview, not just a certificate.
NOT FOR YOU IF
- You want a self-paced course you can binge in a weekend; this is live and cohort-paced.
- You're looking for a credential alone, with no intention of doing the labs.
- You can't attend live or watch the recording within a few days.
- You're already deep into production ML and need advanced or research-level content.
Curriculum
Thirteen weeks plus Demo Day, in a real sequence. Each week names the outcome and the artifact you walk away with. Expand any week for the full session breakdown.
By the end you've built a working AI agent with zero code, then hit its limits and started writing Python to get past them.
n8nGoogle SheetsTelegramPythonBy the end you can load, clean, and visualize a real dataset, and you've already spotted the pattern you'll model in Week 3.
NumPypandasMatplotlibSeabornBy the end you've trained, evaluated, and can explain your first model end to end.
scikit-learnpandasBy the end you have two models packaged into one deployed app, and you understand that deployment is a repeatable process, not a property of the model.
scikit-learnMLflowStreamlitFlaskDockerBy the end you've run an object detection demo and made your first LLM API call, closing the gap between numbers, pixels, and words.
PyTorchYOLOOllamaBy the end you've shipped a real chat UI with tool calling, and you understand what RAG is and why it works.
GradioLangChainChromaDBBy the end you've shipped a RAG app over your own documents and know how and why RAG breaks.
LangChainChromaDBBy the end you've fine-tuned an open-source model and compared it head-to-head against a frontier model.
Hugging FaceLoRAQLoRAGoogle ColabBy the end you've hand-built an agent loop from scratch and deployed your own digital twin.
PythonOpenAI Agents SDKGradioBy the end you've built and deployed a multi-agent research system that's usable outside the classroom.
OpenAI Agents SDKBy the end you've shipped a multi-agent project where a crew of agents with different roles collaborates on real work.
CrewAIBy the end you understand MCP and have started your capstone MVP.
MCPPower BIBy the end your capstone MVP is live in production and every open question is answered.
AWSBy the end you've presented a real, deployed project built in your own domain, from zero code, and you have a portfolio to show for it.
Everything above
What you build
Fourteen weeks, fourteen artifacts. Every week ends with something that runs, in sequence, from a no-code agent in Week 1 to a deployed capstone you present on Demo Day. Step through them.
Week 01 · See It Work, Then Hit the Wall
Three no-code agents, and your first Python chatbot
A portfolio rebalancer, a daily digest agent, and an inquiry triage bot running in n8n, plus a Python chatbot whose if/elif logic you swap for a real LLM call.
- n8n
- Google Sheets
- Telegram
- Python
01 / 14
Who's teaching
Abdul Wahab
AI Engineer · Pécs, Hungary
I've been teaching Python, SQL, machine learning, deep learning, and, more recently, LLMs, RAG, and AI agents online since June 2024. I teach the workflow I actually use: real datasets, the cleaning and debugging nobody puts in the slides, and getting a model out of a notebook and into something that runs.
- 3+ years mentoring experience
- Mentored 150+ students from all around the world
- 1,500+ one-to-one lessons taught
Student reviews
Unedited words from students I've taught one-to-one — in Python, SQL, statistics, machine learning, and AI agents.
Questions
Still have questions? hello@trajectai.org
Cohort 1 starts August 8, 2026.
Basic Python is helpful but not required. We start from the fundamentals.
FREE · Live on Zoom · August 8 – November 7, 2026
Registration closes August 4, 2026