Skip to content
Cloud ForestONLINE ACADEMY
Visit our school

Assessed multi-course certificate · non-accredited

Applied AI Integration

Build AI tools that use sources, ask before making changes and help with text or media. We read your code and explanation. This review does not run your whole app. You need earlier study in databases, secure servers, testing and AI basics.

5 courses. One defined path.

Follow the required courses in order. Audit study is available with Academy membership; assessed enrollment adds evidence review and the program’s progression rules.

  1. Build Your First AI API Integration

    Course 1 of 5 · Advanced · 8 lessons · about 10 h

  2. AI Answers Grounded in Your Sources

    Course 2 of 5 · Advanced · 12 lessons · about 20 h

  3. AI Tools and Approved Workflows

    Course 3 of 5 · Advanced · 9 lessons · about 16 h

  4. Connect Tools with MCP

    Course 4 of 5 · Advanced · 9 lessons · about 12 h

  5. Working with Images, Audio and AI

    Course 5 of 5 · Advanced · 9 lessons · about 12 h

What you’ll work toward

  • FAQ assistant: source and evaluation-design review
  • Grounded answers: source and evidence analysis
  • Approved actions and MCP: boundary source review
  • Editable media assistant: source and review-method assessment
  • Integrate the actual course source/artifacts and explain a documented revision within the stated review limits.

Assessment & pacing

Each assessed lesson opens at least 2 days after the previous lesson is completed. Required assessment checkpoints also need approval before you continue.

Plan for about 12 weeks at 6 study hours per week, based on approximately 70 hours of course study. Allow extra time for revisions.

The minimum assessed practice period is 16 weeks. Studying faster does not bypass lesson waits or required approvals.

The assessed enrollment has a 232-day completion window from its recorded start. The clock pauses while submitted work waits for review. You can request a free extension if you need more time. Membership study remains self-paced while access is active. Paying or marking a lesson complete does not prove proficiency.

How an assessment passes

Each skill is scored from 0 to 4. You need at least 3 on every essential skill and at least 80% overall. Missing evidence must be resolved before approval. Feedback explains what to improve, and you can submit a revised attempt.

To earn the program award, complete the whole course bundle, required quiz passes, approved assignments, final project, and practice period. The school confirms the award separately.

Assessment 1 · code

FAQ assistant: source and evaluation-design review

Make a practice FAQ assistant that cites its sources and checks its answers. Send the code and a plan for checking eight different cases.

Detailed assignment instructions

Send the practice Academy FAQ assistant code, or connected sections that show the whole task. Include the offline option, answer checks, source links, account and usage limits and the mock support draft that needs approval. Plan eight check cases and state the expected results. Follow an unsupported answer and a refused action through the code. We do not make live AI calls or accept a claimed accuracy score as proof. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

See assessment criteria
  • Source and artifact brief (35%) · essential
    The submitted source and eight-case evaluation design cover the actual FAQ-assistant brief.
  • Source and design decisions (30%) · essential
    Validation, source links, access/quota handling and explicit draft confirmation are visible and internally coherent.
  • Source analysis and check design (25%) · essential
    Trace ordinary, unsupported, denied and offline cases through source locations and expected outcomes; identify what future execution would need to verify.
  • Explanation, attribution and revision (10%)
    Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Assessment 2 · code

Grounded answers: source and evidence analysis

Make an assistant that answers from documents you are allowed to use. Send the code, a sample document and examples that show when it should say it cannot answer.

Detailed assignment instructions

Send the document assistant code and a readable sample source you are allowed to use. Show how the code limits source access, links answers to sources and handles missing support. Compare sample answer claims with the text you send. Plan checks for a question outside the sources, conflicting sources and a source the user may not read. We review the design and examples; we do not verify live search quality or an accuracy score. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

See assessment criteria
  • Source and artifact brief (35%) · essential
    The source and readable examples address permitted-document retrieval, citations and unsupported-answer handling in the course brief.
  • Source and design decisions (30%) · essential
    Permission, source support and uncertainty decisions are explicit and coherent in the artifacts.
  • Source analysis and check design (25%) · essential
    Compare provided claims with inline source evidence and design missing/conflicting/unauthorized-source cases with expected decisions, without relying on reported runtime success.
  • Explanation, attribution and revision (10%)
    Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Assessment 3 · code

Approved actions and MCP: boundary source review

Make a tool that asks before changing a plan and a service that looks up resources. Send the code and show how each handles requests it should refuse.

Detailed assignment instructions

Send the learning-plan tool that previews a change and waits for approval. Also send the resource-lookup MCP service code and its request rules. Follow a valid lookup, bad input, an unauthorized request and an update with no approval. Point to the code that checks approval for the exact change. Describe checks and expected results. We do not run these tools or certify a live service. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

See assessment criteria
  • Source and artifact brief (35%) · essential
    The source artifacts cover both the actual approved-update assistant and validated MCP lookup brief.
  • Source and design decisions (30%) · essential
    Explicit approval, argument validation and identity/access decisions are visible and coherent at the action boundary.
  • Source analysis and check design (25%) · essential
    Trace invalid, unauthorized and unconfirmed paths and propose negative checks with expected results; do not equate student-reported tests with verified execution.
  • Explanation, attribution and revision (10%)
    Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Assessment 4 · portfolio

Editable media assistant: source and review-method assessment

Make a tool for editing a transcript or image description. Send the code, a readable example and a plan for checking and correcting mistakes.

Detailed assignment instructions

Send the transcript or image-description editor code and design. Include a readable example, a correction you made and rules for using permitted media. Explain how uncertainty is shown. Describe how a qualified person would compare the result with the original media. We read the code, examples and check plan. We do not view inaccessible media or verify transcript or image accuracy. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

See assessment criteria
  • Source and artifact brief (35%) · essential
    The source/design and readable examples address the authored editable-media assistant brief.
  • Source and design decisions (30%) · essential
    Permission, uncertainty, editable corrections and review responsibilities are explicit and internally coherent.
  • Source analysis and check design (25%) · essential
    Provide a source-linked correction and concrete media-comparison method with failure examples; distinguish a proposed review from independently verified media accuracy.
  • Explanation, attribution and revision (10%)
    Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Assessment 5 · capstone · portfolio

AI system: integrated source and evaluation portfolio

Bring your AI projects together in one clear project pack. Send the code, sample sources, check plans and an explanation of a change you made.

Detailed assignment instructions

Bring the FAQ, document-answering, approved-action, MCP and media-editing projects into one project pack. Send connected code sections, permitted sample documents, answer and access checks, approval rules and check plans. Add notes about failures and recovery. Follow a supported answer, unsupported claim, refused action and manual media correction through the design. Show and explain one change. We do not run this system or accept a claimed pass as proof that it is safe. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

See assessment criteria
  • Source and artifact brief (35%) · essential
    The source portfolio integrates all five actual applied-AI course projects and their stated briefs.
  • Source and design decisions (30%) · essential
    Source support, permission, quota, explicit approval, tool-schema and media-correction decisions are coherent across the supplied artifacts.
  • Source analysis and check design (25%) · essential
    Trace supported/unsupported/denied/failure cases to source locations, meaningful expected outcomes and revision evidence; distinguish design review from runtime validation.
  • Explanation, attribution and revision (10%)
    Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

School recognition, clearly defined

A non-accredited Cloud Forest school award based on reviewed evidence. It is not a degree, professional license or guarantee of employment. Martial rank requires the authorized examiner's approval.

Historical completion awards remain part of a student’s record. New assessed recognition requires the program’s reviewed evidence and approval.