Your SAP AI Portfolio — Three Demos an Analytics Consultant Can Show

What you will learn

• Explain why a working demo outweighs a certificate in a first client or hiring conversation, and name the SAP AI credentials that exist (C_AIG, Joule Studio developer, Business AI Platform solution architect)
• Build or fully specify a grounded RAG assistant using the generative AI hub orchestration service, including its fixed module order and the one mandatory module
• Combine a tabular prediction (SAP-RPT or HANA PAL) with an LLM explanation step that never invents a reason the model did not compute
• Scope a narrow, single-tool agent built with Joule Studio's agent builder, naming who executes the tool call and under what identity
• Identify and deliberately demonstrate one failure case per demo, and explain what it proves about grounding, prediction confidence or agent boundaries
• Package a demo into a reusable portfolio artefact (architecture note, recording, build note) without exposing real client or personal data

Module overview

Who this is for. You have covered the SAP AI fundamentals, tried at least one hands-on exercise in the generative AI hub or SAP HANA Cloud's Vector Engine, and worked through what a scoped SAP AI engagement looks like (“Building the SAP AI Business Case — Value Hypothesis, AI Units and Proof Before Production” to “Scoping and Pricing SAP AI Engagements”). The gap left is not knowledge — it is proof. A hiring manager, a headhunter or a client sponsor who reads "SAP AI" on a CV asks one question before any other: "show me something you built." This module does not teach a new SAP feature. It turns three things you already partly know into three small, honest demos you can run from a laptop in under 20 min each, on data you are allowed to show. The credential that sits next to the portfolio is SAP's generative AI developer exam, covered in section 1.

Prerequisites

• Completion of AI & LLM Fundamentals for SAP Consultants or equivalent working knowledge of tokens, embeddings, RAG and agents
• Familiarity with at least one of: the generative AI hub orchestration service, the HANA Cloud Vector Engine, or SAP-RPT/PAL (“SAP HANA PAL Hands-On — Classification, Clustering and Association with UnifiedClassification” to “hana-ml Hands-On — DataFrame, ConnectionContext and ModelStorage in SAP HANA Cloud”)
• Optional: a generative AI hub trial tenant or a BTP account with AI Core provisioned, for the hands-on variant of E1

Outcomes

• Show a working or fully specified grounded assistant and explain, on request, its orchestration order and its one deliberate failure case.
• Present a tabular prediction with a plain-language explanation that a business user could act on, without overstating model confidence.
• Describe a narrow agent design, including the executing identity and the human-in-the-loop step, in terms a non-technical sponsor can follow.
• Produce three portfolio artefacts (architecture note, short recording, build note) that expose no real client or personal data.