01 / Independent Advisory, Zurich

Internal Audit & AI Education for Risk-Aware Organizations and Ambitious Professionals

I help organizations strengthen internal controls and use AI safely, and help individual professionals use AI tools more effectively in their own work, so you can innovate without losing control.

Based in Zurich, working with clients across Switzerland and Europe.

02

Services

Corporate Services

For organizations: independent assurance and AI-ready teams.

Internal Audit & Controls

Independent assurance, clear insight for management and the board, and stronger control environments grounded in proven Big 4 methodology.

  • Internal audit engagements (process, function, entity and project reviews)
  • Internal control system (ICS) design and effectiveness assessments
  • SOX / ICFR readiness and remediation support
  • Compliance & risk assessments (policies, procedures, governance structures)
  • Third-party / vendor risk management programs
Learn More About Internal Audit & Controls

AI Education for Teams

Most organizations already have AI in daily use, with no training and no rules. I design AI literacy programs that are practical, safe, and built around how your teams actually work.

  • AI literacy programs aligned with your tools, policies and risk appetite
  • Role-based workshops for managers, knowledge workers, and audit/risk/compliance teams
  • Hands-on sessions built around real scenarios from your organization
  • A simple "AI usage rules" playbook employees can actually follow
Learn More About AI Education for Teams
Individual Services

For professionals: think, learn, and work better with AI.

AI Coaching for Individuals

Most people are already using AI without any real training. I help individual professionals build a clear, safe, and genuinely useful way of working with LLMs and agents.

  • How LLMs work: what they can and cannot do, and how to prompt them effectively
  • How to use an AI agent: agentic workflows, when to use an agent vs. a simple chat, and how to keep human control
  • How to secure your data before using AI: what never to paste into a tool, how to choose tools, how to avoid data leaks
  • Building your own library of prompts, workflows and routines for emails, analysis, drafting, learning and more

Formats: 1:1 sessions or small groups; a personal AI setup with recommended tools, workflows and safety rules for your context.

Learn More About Individual Coaching
03

Services in detail

Corporate Services, in detail

Independent assurance, controls maturity, and team-wide AI literacy.

Internal Audit & Controls

Who it's for
Mid-size and large organizations, fast-growing companies, regulated or risk-sensitive environments.
Types of engagements
Support and execution of the audit plan; targeted audits on specific processes, entities or projects; internal control diagnostics and remediation roadmaps; pre-IPO or pre-transaction controls readiness.
Engagement model
Project-based missions, or co-sourcing and advisory alongside internal teams.
Background
Trained at KPMG (Audit Supervisor, La Défense, France), executing statutory and consolidated financial statement audits and advisory assignments for major energy-sector clients including EDF, Veolia Environnement and Dalkia, followed by hands-on internal audit and risk leadership roles inside major international groups including Holcim and Amcor.

AI Education for Teams

Outcomes
Baseline AI literacy across the organization, safer and more consistent use of AI tools, real productivity gains in day-to-day tasks.
Formats
Introductory sessions on AI basics, LLMs, risks and opportunities; deep-dive workshops on Using AI in your daily work; dedicated sessions for internal audit, risk and compliance teams on reviewing AI usage.
Deliverables
Slides, handouts, an AI usage guidelines one-pager, optional follow-up clinics or Q&A.
Individual Services, in detail

A clear, safe, and genuinely useful way of working with AI.

AI Coaching for Individuals

Outcomes
A clear mental model of LLMs and agents; a personal library of prompts and workflows for your work and life; confidence using AI without feeling behind.
Formats
1:1 coaching or small groups, structured around your real tasks and goals.

What a typical engagement covers

  1. 01Map your current AI use, your real tasks, and the data you handle.
  2. 02Pick the right tools for your context and budget, with clear safety rules.
  3. 03Build a starter prompt and workflow library for your recurring tasks.
  4. 04Practice with agents on a real project, with human checkpoints.
  5. 05Set a simple routine so you keep improving after we stop working together.
04

Approach

Applies across both Corporate and Individual services, as one shared process.

01

Understand your context

Business model or personal workflow, regulatory context, existing controls and tools, current AI use.

02

Identify risks and opportunities

Where controls quietly break, where AI could help and where it could hurt.

03

Co-design practical solutions

Audit plans, ICS improvements, AI training programs, usage guidelines, or personal AI routines.

04

Support adoption

Review, Q&A, refinement, and transfer of skills so clients can continue without heavy dependency.

Whether you're an organization or an individual, the process is the same: understand, identify, co-design, and support adoption. I bring audit rigor into AI education, and AI literacy into audit conversations.

05

Case examples

All cases below are Corporate Services engagements.

Case · International NGO

Governance & Compliance Framework for a Sensitive-Service Non-Profit

About

A Swiss-based non-profit running digital support services for vulnerable populations, handling highly sensitive personal data and working with institutional donors and public officials under multi-year grant agreements.

Problem

A major donor agreement carried a material breach clause tied to anti-corruption compliance, with no written policy in place; in parallel, the organization processed high-risk personal data under both Swiss FADP and EU GDPR with no documented governance, exposing leadership to personal liability; restricted donor funds also needed clean separation and audit-ready tracking through a multi-year audit window.

Approach

Drafted and implemented a zero-tolerance Anti-Bribery & Corruption Policy (covering public officials, facilitation payments, gifts and hospitality thresholds, and a reporting process), a Data Protection Policy aligned with Swiss FADP and EU GDPR (including a 72-hour breach notification protocol), and a restricted-fund tracking process (dedicated project coding, expense tagging, and a standing audit folder) to meet donor reporting obligations.

Result

Documented, board-ready governance covering anti-corruption and data protection risk; a defensible compliance position ahead of donor audits; and a clear, repeatable process for tracking restricted funds tied to institutional grants.

Case · Holcim Group

Internal Control & Compliance Across Business Units

Problem

Fragmented controls and compliance exposure across a multinational footprint, with limited shared risk language between entities and service centers.

Approach

As Senior Internal Auditor, ran targeted audit missions mapped to FCPA, UK Bribery Act and Sapin II requirements, coordinated a cross-border data protection review with the Group Data Protection Officer, and helped standardize the rollout of Alteryx Analytics for supplier master data reviews.

Result

Stronger, harmonized compliance coverage across business units and clearer assurance for the Group on anti-corruption and data protection risk.

Case · Amcor Group

SOX Implementation & Enterprise AI Governance

Problem

A global SOX program required tight risk assessment, control design and stakeholder alignment, while informal AI usage was emerging across teams with no governance, policy, or enterprise-wide use rules in place.

Approach

As Audit & Risk Manager, led risk assessment, control design and stakeholder alignment for SOX implementation; built analytics solutions for journal entry testing and anomaly detection; and designed and rolled out an AI governance framework defining acceptable enterprise-wide AI use, alongside a third-party risk management program covering 25 vendors.

Result

Full SOX compliance achieved, a 40% reduction in time spent on journal entry testing and anomaly detection through the analytics solutions built, and a board-ready AI use policy in place ahead of most peers in the sector.

06

About Constantin Roche

01

Experience

I have 10+ years of experience in internal audit, internal control, risk management and compliance. I trained at KPMG as an Audit Supervisor, auditing major energy-sector groups including EDF, Veolia Environnement and Dalkia, before moving into internal audit and risk leadership roles at Holcim and Amcor.

I've worked across audit, transformation assurance and governance, from SOX implementation and ICS design to compliance reviews (FCPA, UKBA, Sapin II) and data protection reviews. I've seen how organizations really work from the inside: where controls quietly fail, where risk is misunderstood, and where processes can be simplified without losing assurance.

02

AI in my own work

I use modern AI tools (LLMs and agents) daily in my own work, and I designed enterprise AI governance frameworks before most organizations had one. I've spent serious time learning what these tools do well, where they fail, and how to use them without leaking data or breaking process.

This is the perspective I bring to AI education: practical, safe usage rooted in real audit and governance experience, not theory.

03

How I communicate

I adapt to your use case and your needs. Executive committee, internal audit function, or a team that has never opened an LLM before, I adjust the language, depth and format so people stay engaged and actually learn.

Direct, calm and hands-on. I explain my reasoning, transfer skills, and leave clients better equipped than I found them.

07

Contact

Replies typically within two business days.