ai-agents

Document Summarisation Pipeline

Create a pipeline that reads uploaded documents, extracts key insights, and sends a structured summary.

deliverables

rubric areas

Intermediate

level

5-7 hrs

time

Our context

Local problems

Datasets, projects, and examples stay close to the markets learners know.

Need, not hype

Build what helps

Tools are taught through decisions, systems, and practical output.

Today

Mentor review queue

12 pending

Submissions

Healthy

At risk

LangChain + EmailAutomation

What you learn

Build a document intelligence flow that saves manual reading time.

Deliverables

Pipeline designSummary outputError handling notes

Scoring rubric

Extraction qualitySummary clarityAutomation flowReliability

Skills practiced

PDF parsingPrompt chainingEmail automationQuality checks

Mentors review the submission for technical accuracy, business reasoning, and manager-ready communication.

How submissions are judged

Data source: Sample PDF documents

Accuracy of the analysis and any calculations.

Whether the recommendation is specific and useful.

How clearly the learner explains the decision.

Whether the output feels ready for a hiring manager.

Score bands

90-100: portfolio-ready

75-89: strong with revisions

60-74: needs coaching

Submit

Send your work for review.

Submit a shareable artifact link so a mentor can review your workbook, dashboard, notebook, or written recommendation.

Related proof

Strong submissions surface here

See the standard learners work toward: clear analysis, practical recommendations, and portfolio-ready delivery.

Product Growth Funnel Analysis

Alice Mwangi

93%

A funnel readout with cohort breakdowns and a practical activation experiment.

See showcase

SME Collections Priority Model

Kevin M.

90%

A prioritized collections list with risk bands and expected cash recovery.

See showcase

Customer Support Quality Scorecard

Naomi A.

88%

A support dashboard that connects response time, quality, and coaching actions.

See showcase

Open showcase