Built for AI Adoption Manager review

Personal application project

A recruiter-ready CV build, explained end to end

I saw the AI Adoption Manager role, decided not to send a generic PDF, and built a browser-first CV workflow that turns one verified Marcel Kenner profile into role-specific previews and production-ready PDF exports.

AI Adoption Manager

AI adoption leadership and enablement roles

The clearest leadership-facing framing, with the strongest narrative for adoption, enablement, and stakeholder communication.

View online
Listen to narration

No time to read? Listen to the CV instead.

Download PDF
Best use case
Default recruiter download for AI Adoption Manager and AI enablement applications.
Paper
A4 Portrait · 210 x 297 mm
Language
English

Marcel Kenner

System & Business Analysis | Enterprise Platforms

  • Poznań, Wielkopolskie, Poland
Marcel Kenner headshot

Professional summary

Business and System Analyst with 5+ years of experience translating complex business needs into clear, implementable solutions across discovery, design, delivery, rollout and support. Experienced in insurance, ESG, customer portals and enterprise platforms. Supports AI enablement through AI assistants, RAG-based solutions and business-facing workflow improvement. Combines business analysis, system integration thinking and multilingual stakeholder communication.

Core competencies

  • AI adoption and enablement
  • AI use case discovery
  • Business analysis
  • Requirements engineering
  • Workflow and process improvement
  • System integration design
  • API-based solution design
  • Stakeholder management
  • Backlog and delivery support
  • Process and knowledge documentation
  • UAT and rollout support
  • Multilingual communication

Professional experience

System / Business Analyst

Insurance technology organisation delivering bancassurance, life insurance, ESG and customer portal initiatives, alongside internal AI enablement work.

  • Lead end-to-end analysis for insurance and internal AI initiatives, translating business goals into scope, user stories, functional specifications and test scenarios.
  • Identify AI-supported workflow opportunities and shape practical internal use cases with measurable operational value.
  • Define use cases for internal AI assistants and co-design RAG-style flows across distributed data sources.
  • Design API-based integrations between core insurance systems, front-end applications and supporting services, covering data contracts, business rules and edge cases.
  • Partner with developers, architects, UX, QA and product owners to validate solutions, support UAT and rollout readiness, and keep delivery aligned with business and regulatory constraints.

Junior System / Business Analyst

  • Translated business goals into Jira stories, acceptance criteria and functional detail for bancassurance feature teams.
  • Maintained requirements, process flows and decision context in Confluence so delivery teams had a clear, current source of truth.
  • Worked with developers, UX designers and architects to refine solution concepts, remove ambiguity and support rollouts.
  • Led refinement sessions, supported testing and gathered stakeholder feedback to confirm delivered functionality matched expected outcomes.

IT Service Desk Analyst

  • Supported English-, Polish- and German-speaking users in a 24/7 enterprise environment, handling L1/L2 troubleshooting and escalation.
  • Worked across Microsoft enterprise environments including Windows Server, Exchange, Office 365, Active Directory and Azure Active Directory administration.
  • Built a structured knowledge base from scratch, documenting troubleshooting paths and reusable resolution methods.

IT Analyst (First and Second Line Support)

Enterprise support delivery for Airbus Commercial and ITERGO service operations.

  • Supported Airbus Commercial and ITERGO users across first- and second-line service desks, providing remote troubleshooting, ticket routing and resolver-group coordination.
  • Supported the Berlin-to-Poznań transition through knowledge transfer, analyst enablement and training for new team members.
  • Worked with ServiceNow and BMC Remedy while creating knowledge articles and resolving escalated tickets.

Languages

  • English - Fluent
  • German - Fluent
  • Polish - Fluent

Technical skills

  • AI adoption and AI-supported workflows: AI use case identification · Internal AI enablement · AI assistants · RAG-style workflows · Context-aware knowledge support
  • Analysis and delivery: Requirements analysis · Jira epics and stories · Acceptance criteria · Backlog support · Functional specifications · UAT support
  • Systems and documentation: API integration design · Data contracts · Confluence · Process documentation · System documentation

What This Demonstrates

What this project shows in practice

Opportunity recognition

Turning a job application into a product decision

The project started from a concrete application goal, then turned that goal into a useful artifact instead of another static PDF submission.

Tooling judgment

Choosing a stack that fits the output

The HTML/CSS-first plus Playwright route was chosen because it gives tighter control over print CSS, fonts, spacing, and browser-to-PDF fidelity.

Execution discipline

Building from plan to renderer to export

The implementation followed a documented sequence: scope and docs first, then typed data, shared rendering, print routing, PDF delivery, and verification.

Delivery proof

Showing that the output actually works

The final result is recruiter-usable and technically defensible: live preview, print routes, downloadable PDFs, and checks that prove the pipeline holds together.

Why I Built This

Why I built this

The trigger

After seeing the AI Adoption Manager post, the goal was immediate: apply for the role, but do it with something more deliberate than sending another standard PDF. The project needed to feel like a useful recruiting artifact and also demonstrate the kind of AI workflow judgment the role asks for.

The build decision

The decision was to build a CV PDF renderer with Codex, because it was already part of the day-to-day toolchain. The first research question was not visual. It was technical: what stack gives the most control over browser-to-PDF fidelity. The answer was an HTML/CSS-first CV template rendered to PDF with headless Chromium via Playwright, because that keeps print CSS, fonts, spacing, and browser preview behavior under one system.

The execution path

With that direction set, the requirements were broken into focused docs, then into an execution plan and roadmap, and the first pass was followed by a back-and-forth refinement cycle to smooth out errors, layout drift, and misalignments.

How I Built It

Roadmap, implementation phases, and delivery timeline

Total build time

5 hours 9 minutes from Phase 0 completion at 18:49Z to Phase 10 verification at 23:58Z on 2026-03-13.

Implementation window

One focused build sequence on 2026-03-13, followed by later iteration to smooth layout and content details.

Phase 0-2

Completed on 2026-03-13 between 18:49Z and 19:02Z.

Foundation, data model, and source content

What

Locked the first release scope, replaced the starter shell, defined the paper and document types, and created the initial Marcel profile as typed repository data.

How

Followed the roadmap sequence: foundation freeze, app skeleton and tokens, then domain model and sample content before any PDF work started.

Phase 3-4

Completed on 2026-03-13 at 20:24Z and 21:32Z.

Shared renderer and browser workspace

What

Built the reusable CV document component and wrapped it in the homepage workspace with version selection and browser preview.

How

Implemented the shared markup first, then added the showcase route so the document could be reviewed in the browser before print and export paths were introduced.

Phase 5-6

Completed on 2026-03-13 at 21:58Z and 22:14Z.

Print route and paper refinement

What

Added dedicated print routes for A4 and Letter, then tuned spacing and vertical rhythm so the shorter Letter page still looked deliberate.

How

Kept the markup shared and moved paper differences into explicit layout tokens instead of creating separate document branches.

Phase 7-8

Completed on 2026-03-13 at 23:05Z and 23:24Z.

Playwright export and static PDF delivery

What

Added the Playwright PDF renderer, then switched delivery to pre-generated static PDF assets so the site can deploy on Cloudflare Pages without a Node runtime.

How

Opened the internal print route in headless Chromium, waited for print readiness and loaded fonts, called `page.pdf()` with explicit A4 or Letter settings, and wrote the finished files into the static asset tree.

Phase 9-10

Completed on 2026-03-13 at 23:42Z and 23:58Z.

Test harness and final verification

What

Added PDF smoke tests, production-style app startup for Playwright, final documentation updates, and manual preview-versus-export checks.

How

Used request-level end-to-end tests for the PDF contract, then closed the loop with lint, unit tests, build verification, and visual comparison of print routes against exported PDFs.

What We Fixed

Problems we hit and how they were corrected

Fix 01

Random page breaks in PDF output

Problem

The first print/export pass depended too much on normal browser flow, so long sections could break at awkward points once the document crossed a page boundary. The result looked unpredictable instead of intentionally designed.

Fix

The first correction was to measure rendered sections on the print route and force a page break before a section when its heading plus first printable unit would overflow the remaining page space. That moved the decision from passive CSS flow to an explicit pagination plan.

Result

This stabilized the biggest errors, but it also exposed a second issue: moving whole sections could leave a large empty area at the bottom of page 1 and a heavy content block at the top of page 2.

Fix 02

Large blank areas after section-level pagination

Problem

Section-level planning was still too coarse for dense sections such as professional experience. If an entire section had to move, the previous page could end with a lot of wasted space.

Fix

The renderer was refined to mark printable subsection units with `data-cv-unit`, then the planner started paging by section heading plus units instead of treating the whole section as one rectangle. In `Professional experience`, each role is now treated as one printable unit, so roles can move cleanly between pages without splitting the whole section too early.

Result

Pagination became much more intentional. The print route now measures sections and subsection units, applies the plan to the DOM, shows continuation headings where needed, and only then marks the route ready for Playwright PDF export.