See
A visual lesson introduces the system, the decision being made, and the failure mode it prevents.
AI Engineering · complete syllabus
A curriculum with
a reason for every step.
Sixteen guided hours move from model behavior to a production-minded AI system. Each module gives Mira one new capability, then connects it to a lab, a test, and a reusable reference.
The learning method
Understanding should survive
after the slide closes.
The course deliberately combines explanation, worked examples, practice, feedback, and later retrieval. That structure follows established learning-science guidance—not a claim that watching more content creates expertise.
Build
A guided lab turns the concept into working behavior inside Mira, the course's evolving AI employee.
Test
Edge cases, evaluation criteria, and security checks reveal whether the behavior actually works.
Retrieve
A concise field guide and connected article make the pattern findable after the lesson is over.
Eight connected modules · 16 hours
From first prompt
to operating system.
011.5 hours
Capability
AI fundamentals & prompting
Understand probabilistic output, tokens, context limits, and the five-part prompt structure used throughout the course.
Mira can communicate with a defined role, task, constraints, and output shape.Guided practice
Compare prompt structures and convert an unstructured response into a validated result.
- Time
- 1.5 hours
- Leaves you able to
- Mira can communicate with a defined role, task, constraints, and output shape.
022 hours
Capability
Agents & tool use
Move from a single model response to a controlled loop that can select tools, act, observe results, and stop safely.
Mira can choose a permitted tool and use its result without seeing its credentials.Guided practice
Design a minimal tool set, schemas, permission boundaries, and failure behavior.
- Time
- 2 hours
- Leaves you able to
- Mira can choose a permitted tool and use its result without seeing its credentials.
032 hours
Capability
Building an agent
Compose routing, chaining, decomposition, and evaluator patterns into a small agent that is easy to inspect.
Mira can break a goal into bounded steps and explain what happened.Guided practice
Build the travel-agent core, then test tool errors, incomplete inputs, and unsupported requests.
- Time
- 2 hours
- Leaves you able to
- Mira can break a goal into bounded steps and explain what happened.
042 hours
Capability
Memory & context
Separate working context, session state, durable preferences, and source-of-truth data instead of calling everything memory.
Mira can remember an approved preference without leaking it into another user’s session.Guided practice
Add explicit save/retrieve behavior and verify isolation across sessions and users.
- Time
- 2 hours
- Leaves you able to
- Mira can remember an approved preference without leaking it into another user’s session.
052.5 hours
Capability
Knowledge & retrieval (RAG)
Build ingestion, chunking, embeddings, retrieval, grounding, and citations as one measurable system.
Mira can answer from approved documents, cite evidence, and abstain when evidence is weak.Guided practice
Create a small knowledge base, inspect retrieved chunks, and test unanswerable questions.
- Time
- 2.5 hours
- Leaves you able to
- Mira can answer from approved documents, cite evidence, and abstain when evidence is weak.
062 hours
Capability
Production engineering
Treat cost, latency, reliability, observability, and deployment as architecture—not cleanup after the demo.
Mira can operate within a latency target, cost budget, and defined fallback path.Guided practice
Measure a baseline, remove unnecessary context, add caching, and compare before/after behavior.
- Time
- 2 hours
- Leaves you able to
- Mira can operate within a latency target, cost budget, and defined fallback path.
072 hours
Capability
Evaluation & security
Test task quality and system safety with datasets, thresholds, permissions, adversarial cases, and human escalation.
Mira can refuse unsafe actions, protect secrets, and surface uncertain decisions to a person.Guided practice
Create quality, injection, exfiltration, excessive-agency, and escalation tests.
- Time
- 2 hours
- Leaves you able to
- Mira can refuse unsafe actions, protect secrets, and surface uncertain decisions to a person.
082 hours
Capability
Capstone & review
Combine goals, tools, memory, knowledge, guardrails, evaluation, and operations into one defensible system.
Mira can complete a bounded workflow and show the evidence needed to review it.Guided practice
Deliver an architecture brief, working flow, evaluation results, threat review, and operating plan.
- Time
- 2 hours
- Leaves you able to
- Mira can complete a bounded workflow and show the evidence needed to review it.
What learners leave with
Not a folder of prompts.
A defensible system.
- Explain when an agent is appropriate—and when deterministic software is safer.
- Design tool schemas, permission boundaries, stop conditions, and human approvals.
- Build and evaluate retrieval separately from answer generation.
- Scope memory by user, purpose, freshness, correction, and deletion.
- Measure quality, latency, cost, reliability, and security before deployment.
- Present a capstone with architecture, evidence, failure handling, and an operating plan.
A factual comparison
Choose a learning system,
not a content count.
No curriculum is universally better. This one is designed for learners who want connected engineering practice rather than disconnected demonstrations.
Course characteristicCommon content-first formatMotyWait learning path
StructureIndependent topics or tool toursOne system that evolves across eight modules
PracticeWatch, copy, continueWorked example → guided lab → edge cases
EvaluationDoes the demo run?Task quality, retrieval, reliability, cost, latency, and safety
ProductionDeployment as a final chapterPermissions, observability, failure handling, and operations throughout
After the lessonReturn to a long recordingField guide, searchable articles, and module connections
Why this design: the U.S. Department of Education’s What Works Clearinghouse recommends spacing learning over time, interleaving worked examples with problem solving, combining graphics with verbal explanations, using active retrieval, and asking deep explanatory questions. Read the evidence guide ↗
Course 01 · coming soon
See the whole path.
Then decide if it fits.
The complete teaching deck and answer keys stay private until every learning connection and lab has been reviewed.