AI Business Professional study plan

Define machine learning, deep learning, and generative AI in your own words and distinguish each from rule-based automation

Classify ten real workplace tools as traditional software, ML-driven, or generative AI and justify each label

Explain how a model is trained, evaluated, and deployed using a single concrete business example

Identify the limits of LLMs (hallucination, context windows, knowledge cutoffs) and describe one business risk each creates

Write prompts that specify role, task, context, constraints, and output format

Revise five weak prompts into structured prompts and record what changed in the output

Apply at least three iteration techniques (refinement, examples, chain-of-thought requests) to improve a single output across multiple passes

Diagnose common failure modes such as vague instructions, conflicting constraints, and missing context, and correct each

Map one real business workflow end to end and mark where AI could assist, automate, or should stay human

Draft prompts that support a specific workflow step such as summarizing, drafting, classifying, or extracting

Define human review checkpoints and escalation criteria for an AI-assisted process

Estimate time saved or error reduction for the redesigned workflow using stated assumptions

Distinguish structured, semi-structured, and unstructured data and give a business example of each

Identify the data quality dimensions (accuracy, completeness, consistency, timeliness) and audit a small dataset against them

Explain how training data bias and labeling choices affect model output

Describe basic data governance practices including ownership, lineage, and retention

Identify personal and sensitive data categories and explain why each matters for AI use

Apply a privacy-by-design checklist to a proposed AI use case and flag the highest-risk elements

Explain prompt injection, data leakage, and model output misuse, and describe one mitigation for each

Draft responsible-use guidelines covering transparency, human oversight, and acceptable use

Compare build, buy, and fine-tune versus prompt-only approaches for a given business need and recommend one

Define evaluation criteria for an AI solution and design a simple test set to check output quality

Explain how retrieval-augmented generation differs from relying on model memory and when each is appropriate

Revisit Week 1 concepts and correct any misconceptions surfaced by later material

Build a cost-benefit estimate for one AI initiative including labor, tooling, and expected return

Identify organizational readiness factors such as skills, process maturity, and change management needs

Outline a phased adoption roadmap with quick wins and longer-term capabilities

Frame AI value in business terms such as cycle time, quality, and capacity rather than technical metrics alone

Synthesize all six domains into a single integrated AI use case proposal

Self-assess weak areas by domain and revisit the two lowest-scoring topics

Practice explaining AI concepts to a non-technical stakeholder in plain language

Complete a timed review of key terms, frameworks, and decision criteria across all domains