# AI Workshop Switzerland — Full Program Curriculum # https://aiworkshop.ch/llms-full.txt # This document contains the full detailed curriculums and use-cases for all AI Workshop Switzerland programs. # Purpose: To provide Retrieval-Augmented Generation (RAG) context for LLMs. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PROGRAM: Microsoft Copilot & AI Training for HR Teams ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Reduce shared inbox overload and standardize HR responses using Microsoft 365 Copilot. ## Modules: ### Module 1 - HR-Safe AI & Copilot Foundations - What Copilot can/can't do in HR - HR-safe prompting rules (red/yellow/green) - How to review outputs and stay compliant *Outcome:* Your team uses AI confidently, safely, and responsibly. ### Module 2 - Outlook Copilot for Inbox Triage & Faster Replies - Summarize long email threads into actions - Extract the real request + missing details checklist - Draft replies with consistent tone and clarity - Build standard templates for recurring topics *Outcome:* Faster replies, better consistency, less inbox fatigue. ### Module 3 - Employee Questions System (Knowledge Base Design) - FAQ structure, categories, tags, and ownership - How to write answers employees understand - Link answers to source-of-truth documents *Outcome:* Fewer repeat questions and improved HR scalability. ### Module 4 - Word Copilot for Policy Simplification + Multilingual Clarity - Rewrite policy text into employee-friendly language - Create short versions for quick replies - Translate while preserving meaning and tone *Outcome:* Less confusion and fewer back-and-forth messages. ### Module 5 - Teams Copilot for Meetings, Actions & Follow-Up - Capture decisions and next steps automatically - Turn meeting notes into action lists and follow-ups *Outcome:* Better alignment and fewer lost decisions. ## Frequently Asked Questions: Q: Is this training only for Microsoft Copilot? A: Yes. This program is designed specifically for Microsoft 365 Copilot and fits organisations where ChatGPT is restricted for privacy reasons. Q: How does this workshop comply with Swiss Data Protection (FADP)? A: We align our workflows with Swiss FADP (nLPD) requirements, focusing on anonymization, local data processing in Microsoft 365 Switzerland, and strict 'human-in-the-loop' verification protocols. Q: Can you deliver this on-site in Switzerland? A: Yes. On-site delivery is available across Switzerland (Zurich, Geneva, Lausanne, Zug, etc.) and is recommended for the best hands-on team interaction. Q: Is it safe for sensitive HR data? A: We teach HR-safe prompting, compliance guardrails, and human review workflows so your team stays in control without exposing PII to external models. Q: What specific HR deliverables will my team have after 1 day? A: Your team will leave with a verified HR Prompt Library, standard reply templates for your shared inbox, and a 30-day adoption plan tailored to your specific tools. Q: Do participants need technical or coding skills? A: No. The program is designed for HR professionals and focuses on real daily tasks like email triage, policy simplification, and drafting. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PROGRAM: AI Training for Finance Professionals ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Switzerland's finance sector is under pressure to move faster, report more clearly, and serve clients better - all with the same or fewer resources. AI can help, but finance teams need a practical approach that fits audit-ready standards, regulatory constraints, and the judgment-heavy nature of financial work. This workshop is built for that reality. ## Modules: ### Module 1 - AI Foundations for Finance Professionals - How AI models work and where they introduce risk in finance - Data privacy, confidentiality, and what not to share with AI tools - Output verification habits: when to trust and when to check - Regulatory considerations for AI use in financial services *Outcome:* Your team uses AI with financial-grade accuracy standards and appropriate oversight. ### Module 2 - Financial Reporting & Management Commentary - Structuring monthly and quarterly management reports faster - Drafting financial narratives from structured data - Board pack commentary: clear, concise, decision-ready - Reviewing and improving existing financial reports *Outcome:* Reporting cycles that take days compressed into focused hours. ### Module 3 - Analysis Documentation & Research Summaries - Turning financial model outputs into readable analyst commentary - Synthesizing market research and economic updates into briefings - Competitive intelligence summaries for investment decisions - Peer and sector comparison documents *Outcome:* Analysis outputs that communicate clearly to non-finance stakeholders. ### Module 4 - Client Communication & Relationship Management - Drafting personalized client portfolio updates - Investment proposal and recommendation letter templates - Meeting follow-up notes and action summaries - Clear, accessible financial explanations for private clients *Outcome:* Higher-quality client communication produced in significantly less time. ### Module 5 - Compliance Documentation & Operational Efficiency - Structuring compliance reports and regulatory filings - Policy document drafting and simplification for internal teams - Audit trail documentation patterns - Identifying repetitive finance tasks candidates for AI support *Outcome:* Less time on documentation, more time on value-adding finance work. ## Frequently Asked Questions: Q: Can we use AI with confidential financial data in Switzerland? A: We address this directly. The workshop covers what to share, what not to share, and how to verify that your Microsoft 365 tenant or OpenAI Enterprise environment meets your institution's data residency and confidentiality requirements. Q: Is AI accurate enough for financial reporting and audit trails? A: We teach 'Audit-Aware AI'—where AI is used for structuring and drafting narratives, while calculation and verification remain in your controlled ERP or spreadsheet environment. Every output requires a human review step. Q: Does this work with Microsoft 365 and Excel Copilot? A: Yes. We have a dedicated focus on Microsoft Copilot for finance, covering Excel for data insights and narrative generation, Word for reporting, and Teams for coordination—the standard approved environment for many Swiss institutions. Q: Is this relevant for private banking and wealth management? A: Absolutely. We have a specific track for private banking covering client communication, portfolio update drafting, KYC/AML documentation support, and personalised investment proposal workflows. Q: What ROI can a finance team expect after this training? A: Most teams report a 30–50% reduction in time spent on narrative drafting for monthly and quarterly reports, and measurably higher quality in client correspondence within the first month of adoption. Q: Can this be delivered to our full finance team on-site? A: Yes. On-site delivery is the preferred format. We tailor the content to your institution's approved tools, workflows, and regulatory environment—FINMA, MiFID II, or internal standards. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PROGRAM: AI Training for Legal Professionals ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Legal professionals are under constant pressure to deliver high-quality work faster. AI can help - but only when you understand where it adds value, where the risks lie, and how to use it in a compliant, judgment-led way. This workshop is built specifically for lawyers, legal teams, and compliance professionals who want to work smarter, not just faster. ## Modules: ### Module 1 - AI Foundations for Legal Professionals - How large language models work and where they fail - Professional responsibility and AI: what the rules say - Hallucination risk, verification habits, and output quality control - Confidentiality and data handling (what to share, what not to) *Outcome:* Your team uses AI with legal-grade judgment and appropriate oversight. ### Module 2 - Contract Review & Clause Analysis - First-pass review prompts for NDAs, service agreements, employment contracts - Flagging unusual, onerous, or missing clauses - Side-by-side comparison of contract versions - Generating structured review summaries and risk flags *Outcome:* Faster first-pass review without compromising legal quality. ### Module 3 - Legal Research & Regulatory Monitoring - Synthesizing case law summaries and legislative updates - Building research memos faster with AI-assisted structuring - Monitoring regulatory changes in Swiss and EU law - Cross-referencing guidance with internal policies *Outcome:* Research that used to take days compressed into focused hours. ### Module 4 - Due Diligence & Document Triage - Structuring due diligence checklists with AI assistance - Document categorization and priority flagging - Generating deal summaries and key findings reports - Handling large document sets efficiently *Outcome:* Faster deal timelines and more consistent due diligence outputs. ### Module 5 - Client Communication & Drafting - Drafting clear, plain-language client letters and updates - Legal correspondence templates for common situations - Translating complex legal language for non-lawyer audiences - Maintaining tone of voice and professional standards *Outcome:* Professional client communication produced in a fraction of the time. ## Frequently Asked Questions: Q: Is it safe to use AI with confidential legal documents? A: We cover this in depth. The workshop includes a dedicated legal data handling module covering which tools are appropriate for different client data types, what zero-retention means in practice, and how to protect attorney-client privilege. Q: How do we maintain professional secrecy (Art. 321 SCC) when using AI? A: We detail the technical measures (TOMs) required to use AI without breaching professional secrecy—including private tenant configurations, zero-retention settings, and when only on-premises models are appropriate. Q: Can AI replace legal judgment? A: No—and this program is designed around that principle. AI handles the mechanical, time-consuming parts. The professional judgment and final sign-off stay with the lawyer. Q: Does the program cover Swiss-specific legal research (Fedlex, BGer)? A: Yes. We demonstrate workflows for synthesizing Swiss case law and monitoring Fedlex legislative updates, using AI-assisted research patterns that respect Swiss jurisdiction. Q: Which AI tools do you cover—ChatGPT, Claude, or Copilot? A: We are tool-agnostic and focus on whichever tools your firm has approved—ChatGPT Enterprise, Claude, or Microsoft Copilot—ensuring they fit your security and compliance environment. Q: Is this suitable for both partners and associates? A: Yes. We tailor the content so partners understand governance and ROI, while associates master the daily drafting and research workflows that save the most time. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PROGRAM: AI Training for Marketing Teams ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Marketing teams are expected to produce more - more content, more campaigns, more channels, more personalisation - with the same resources. AI can genuinely help, but only if your team knows which tools to use, how to brief them well, and how to maintain brand quality and consistency. This workshop is built for marketing professionals who want real output improvement, not just a technology overview. ## Modules: ### Module 1 - AI Foundations for Marketers - How generative AI works and where it falls short in marketing - The brief is everything: how good input creates good output - Brand voice, tone, and consistency guardrails for AI use - Copyright, image rights, and disclosure considerations *Outcome:* Your team uses AI confidently while protecting brand quality and compliance. ### Module 2 - Content Creation & Copywriting at Scale - Writing high-converting headlines, body copy, and CTAs with AI - Adapting content across formats: email, social, landing pages, ads - Maintaining brand voice consistency across AI-generated content - Building content brief templates that produce better outputs *Outcome:* More content, faster - without losing the quality your audience expects. ### Module 3 - Campaign Ideation & Strategy - Using AI as a creative thinking partner for campaign concepts - Developing messaging angles for different audience segments - Rapid A/B variation generation for testing - Campaign brief structuring and kickoff documentation *Outcome:* Faster campaign starts with stronger creative foundations. ### Module 4 - Multilingual Marketing for Swiss Teams - Adapting content across DE, FR, IT, and EN without losing meaning - Cultural nuance and localisation beyond translation - SEO keyword alignment across languages - Community management response templates in multiple languages *Outcome:* Consistent, authentic messaging across all of Switzerland's language regions. ### Module 5 - Reporting, Analytics & Audience Insights - Synthesizing performance data into actionable campaign summaries - Writing clear, executive-ready marketing reports with AI assistance - Extracting audience insights from survey and feedback data - Competitive intelligence: tracking and summarizing market activity *Outcome:* Reports that take half a day compressed into focused effort that actually informs decisions. ## Frequently Asked Questions: Q: Can AI handle the multi-language complexity of the Swiss market? A: Absolutely. We focus on 'Localize, Don't Translate' workflows that adapt tone and cultural nuances across EN, DE, FR, and IT, rather than simple literal translation. Q: How do we prevent 'AI-looking' or generic marketing content? A: We teach brand-voice cloning techniques and specific style-guide prompting so the output sounds like your brand, maintaining your unique competitive edge. Q: Can we use our own brand guidelines and campaigns during the workshop? A: Yes, we encourage working with real briefs and real brand guidelines. This ensures the output is immediately applicable to your current marketing goals. Q: Does AI content rank well in Google search for local SEO? A: We cover how to use AI for high-quality, helpful content that meets Google's E-E-A-T criteria, specifically for localized SEO in the Swiss market. Q: Is AI image generation (Midjourney/Canva) included in the session? A: Yes, as a core part of our content track. We show how to generate on-brand visual assets while navigating the legal and copyright landscape of AI art. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PROGRAM: AI for Developers & Engineering Teams ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Most developers already use AI to write code. Few use it well. This 2-day intensive workshop focuses on what actually matters in a professional environment: effective tool mastery across GitHub Copilot, Microsoft Copilot, and Claude Code - combined with governance, security, and the agentic frameworks that let AI handle real multi-step engineering tasks. ## Modules: ### Day 1 - Module 1: AI Tool Mastery (GitHub Copilot, Claude Code, Microsoft Copilot) - How each tool differs and when to use which - Context injection, multi-file awareness, and working with large codebases - Advanced prompting patterns for code generation, refactoring, and debugging - Claude Code for longer reasoning chains and complex tasks *Outcome:* Developers can use all three tools effectively for real engineering work - not just line completions. ### Day 1 - Module 2: Security, Compliance & Code Governance - Common risks: hallucinated dependencies, credential exposure, GPL-tainted code - Internal policy alignment: what AI can and cannot touch in your environment - Code review workflows adapted for AI-assisted development - Practical red/amber/green guardrails for the team *Outcome:* A shared governance standard your team can adopt from day one. ### Day 2 - Module 3: Agentic AI for Developers - What agentic frameworks are and how they work (Copilot Workspace, Claude Agents, custom agents) - Task decomposition: breaking engineering tasks into agent-executable steps - Hands-on: building a simple agentic workflow for a real use case - Limitations, failure modes, and when not to use agents *Outcome:* Participants can design and run basic agentic workflows for engineering tasks. ### Day 2 - Module 4: Team Standards & Structuring AI Use at Scale - Defining AI usage tiers across your org (who uses what, and how) - Prompt libraries and shared conventions for consistency - Measuring the impact of AI tool adoption - Onboarding new developers into an AI-enabled workflow *Outcome:* A team-level structure for sustainable, governed AI-assisted development. ## Frequently Asked Questions: Q: GitHub Copilot vs. Claude Code vs. Cursor: which do you teach? A: We cover the ecosystem as a whole. We focus on mastery of the tool your company uses while demonstrating the unique advantages of reasoning models like Claude 3.5 Sonnet for coding. Q: How do we handle security and IP when generating code with AI? A: We teach 'Secure-by-Design' AI coding, covering context filtering, VPC-only environments, and how to prevent secret leaks or GPL-tainted code from entering your repo. Q: Does this workshop cover AI agents and agentic frameworks? A: Yes. Day 2 is focused on the move from 'autocomplete' to 'agentic.' We show how to build and govern agents that handle multi-step tasks from Jira to PR. Q: Do participants need to be senior developers? A: No. The program is valuable for junior to senior levels. We focus on engineering patterns and the new 'architectural' way of thinking required for AI-assisted development. Q: Is this delivered on-site or remotely for engineering teams? A: Both. While remote works well for distributed engineering teams, we offer on-site 2-day Intensives in Switzerland for high-bandwidth collaborative learning.