Agentic AI Use Case Deep Dive — Half-Day ILT (Follow-up)

About this session

A hands-on follow-up to the 0.5-day Agentic AI Foundations Workshop, for staff who identified a promising use case and want to take it further. Where the Foundations session builds awareness and a first idea, this Deep Dive is about refining and implementing that idea — deeper agentic patterns, a robust agent design, a full build with verification, and a clear path to hand over to the DNA / IT team.

Where Foundations had everyone build the same generic assistant (the Guideline Assistant) and sketch a canvas, this session moves to each participant’s specific use case. Participants have flexibility: continue from the Agent Design Canvas they started in Foundations, or bring a new problem statement from their work. Either way, they leave with a refined design, a working assistant on AgentSea, and an implementation plan. Worked examples by function are provided for anyone who needs a starting point — Vendor Proposal Analysis (Procurement/Finance), Policy & Circular Compliance (Governance), Claims / Leave Processing (HR), and Budget Variance Commentary (Finance).

Learning Outcomes

By the end, participants can:

1. Select the right agentic workflow pattern for their use case from the four core patterns.
2. Write a robust skill / system prompt with guardrails and escalation rules for their use case.
3. Build and iterate a working assistant on AgentSea, including a verification step that catches errors.
4. Design an agentic automation plan — triggers, steps, and human-in-the-loop checkpoints — ready for when AgentSea’s Workflow Builder lands.
5. Complete an Agent Design Canvas and prioritise their use case on value, feasibility and risk.
6. Define measurable success metrics and an implementation path — what to hand to the DNA / IT team.

Agenda

Timings are adjustable. Runs equally well as a morning or afternoon session.

Module Detail

D0 · Welcome + bring your use case (15 min)

• Quick recap of the Foundations essentials and the agent mindset.
• Each participant (or small team) frames their use case — continue from the Agent Design Canvas they started in Foundations, or bring a new problem statement from their work.
• Anyone who prefers a ready-made starting point adopts a worked example (Vendor Proposal, Policy & Circular Compliance, Claims/Leave, or Budget Variance).

D1 · Agentic workflow patterns in depth (40 min) — agentic AI workflow

• The four core patterns explored properly: Chaining · Parallelisation · Routing · Orchestration — with healthcare-admin scenarios.
• The agent loop (Observe → Plan → Act → Reflect) and levels of autonomy revisited at working depth.
• Applied: each participant picks the pattern that best fits their own use case and sketches the flow.

D2 · Build your agent’s skills & instructions (40 min) ⭐ Hands-on — skills

• What an agent’s skills are, and how the system prompt (instructions) steers them — role, output specification, guardrails, escalation, and explicit “must NOT” rules.
• Applied (hands-on): each participant builds their agent — writes the instructions for their use case and configures the skills it needs from what AgentSea offers today (Document Analysis, Document Generation, Document Builder, Outlook), designing for verification and safe defaults (SGD, PII redaction, never-invent-data).
• Roadmap note: authoring custom, reusable skill packages is a concept (Skills Marketplace, subject to platform readiness) — today, skills are configured via instructions + the built-in set.

D3 · Build & verify — capstone (55 min) ⭐ Full hands-on

On current AgentSea (single agent + Document Analysis + Document Generation + system prompt):

1. Deploy the system prompt drafted in D2 for your use case.
2. Upload a document pack; run Document Analysis → structured output.
3. Iterate on a deliberate-error document — the trust-but-verify teaching moment; refine the prompt until the agent catches it.
4. Generate the final output with Document Builder — a brand / template-compliant deliverable for your use case (e.g., a vendor comparison in Excel, a formatted variance memo/PPT, or a policy briefing) — and review quality.

• This is the complete build experience that the Foundations session only previews.

D4 · Agentic Automation — design your automation plan (30 min) — concept + design

The shift from an assistant you invoke to an agent that runs on a trigger or a schedule — the L1 → L2 → L3 autonomy ladder. This is the “end-to-end automation, not just information retrieval” that staff have asked for.

• Three questions for any automation: What triggers it? · What runs? · Where does a human stay in the loop?
• How it will land on AgentSea: a visual Workflow Builder (manual / scheduled / event-driven triggers, conditional logic, parallel paths, human-in-the-loop approval nodes, orchestration across Knowledge Base Spaces and M365/SharePoint) — a future capability, so this module is design-only (no hands-on build in this session).
• Applied — the deliverable: each participant designs the “automation view” of their agent and captures it as an automation plan (Markdown) — the trigger, the steps, the approval/human checkpoints, and the fallback — ready to hand to the DNA / IT team when the capability lands. This plan feeds the implementation section of the canvas.

D5 · Use-case refinement — Agent Design Canvas (35 min) ⭐ Working session

• Complete the full Agent Design Canvas for the use case: mission, workflow pattern, data, skills, guardrails & escalation, autonomy level, automation trigger, success metrics, and first milestone.
• Prioritise on value × feasibility × risk (and responsible-AI / PDPA considerations).
• Define measurable success metrics and the implementation path — what is ready now on AgentSea, and what to hand to the DNA / IT team (connectors, knowledge spaces, orchestration and scheduling shown as concepts, subject to platform readiness).

D6 · Showcase & next steps (10 min)

• Rapid share-out of designs; peer and facilitator feedback.
• Each participant leaves with a refined canvas, a working assistant, and a first implementation milestone.
• How to engage the DNA / IT team and the champion network.