Fabian Rohmann-PuschnerFRP DE
AI enablement · AI automation in marketing and sales

From source
to agent.

I bring AI into marketing and sales: I find the tasks where it helps, build the workflows myself and enable teams to work with them. And I work the way I learned as a history student: every claim needs a source.

Ask me See work samples
About me

Ask me whatever you'd like to know.

I'm a B2B marketer with eight years of experience, and since 2025 I have been building AI workflows that run in everyday work: research, reporting, content and data maintenance. What I enjoy most is the moment a team realises that a tedious task now gets done on its own.

Instead of a long text, there's a conversation here. Type your question or pick a suggestion. The answers come from my verified documents.

By design, the chat doesn't name customers, partners or internal business figures of former employers. For anything else, feel free to email me directly.
Fabian Rohmann-Puschneranswers from his documents
Prepared answers
Career

One common thread, five stations.

Making complex things understandable and improving processes systematically. First as a historian, then as an editor and in industrial marketing, today with AI agents. Every station left me with something I use every day.

2009to 2017
UniversityTU Darmstadt

M.A. and B.A. in History

Technical University of Darmstadt
  • Finding, checking and weighing sources against each other
  • Structuring large amounts of material and arguing precisely
Source criticismNo claim without evidence.
2017to 2020 · 3 years
AgencyGiessen

Editorial trainee, then technical editor

transQUER GmbH · content marketing agency
  • Project lead for corporate magazines and content campaigns in German and English
  • Clients including Heraeus, Linde, Bayer, Schott and Wacker
  • Advised clients in pitches on social media and influencer marketing
EditorialExplaining technology so people can work with it.
2021to 2025 · 4 years
IndustryReiskirchen

Marketing Specialist

Weiss Klimatechnik · cleanrooms, healthcare, data centres
  • Technical articles, white papers, case studies and LinkedIn for each business unit
  • Product launches with sales and training material
  • Lead campaigns, SEO and SEA, international localisation
B2BSelling complex products without overselling.
202509 to 11/2025
Mid-sized businessHungen

E-Commerce Manager

M24 GmbH · building materials distributor, around 80 employees
  • Developed a company-wide AI roadmap and implemented the first automations
  • ERP, CRM and web shop in Microsoft Dynamics 365 BC, master data governance
  • Position cut for operational reasons, explicitly not performance-related according to my reference
AI adoptionAI starts with the process, not the tool.
202604 to 10/2026
Software · channel salesDACH

Marketing Specialist DACH

Wildix · communications software with AI, sold 100% through partners
  • AI partner activation programme: recruiting workshops, multi-stage onboarding for AI products
  • Built 14 AI workflows for research, reporting, content and data maintenance
  • Localised campaigns with partners for medical practices, law firms and manufacturing
Sources requiredEvery agent backs up what it claims.
Today · How I work

Sources required, around the clock.

This is what a Monday morning in my AI operating system looks like. The workflows run on schedule before the team starts. By 8:30, reports and drafts are ready, and none of it goes out without my approval.

Monday, 07:00 · Scheduled runsrunning
  1. Campaign dashboard regeneratedfive CRM objects, one fixed addressOK
  2. Last week's support tickets clusteredreport for support, topics for marketingOK
  3. Partner posts reviewedrepost drafts in the content pipelineOK
  4. Partner lead handover reportstalled leads flagged, email draft readyOK
  5. Market and competitor reportactions for four departments, every claim sourcedOK
  6. I review and approvethe agent prepares, the human decidesAPPROVAL
Scopesince 06/2026
14
scheduled workflows in day-to-day work
>15
self-written skills and checklists
Time savedestimate
45–50hrs/month
a conservative estimate, not measured. Some of these tasks would not have been done regularly at all without automation.
Rules in every workflowGovernance
  • Draft, don't sendAgents never send, post or delete anything themselves.
  • Read-only where possibleAgents leave the CRM unchanged.
  • A source for every claimData gaps are flagged, not papered over.
  • SafeguardedExternal input is protected against prompt injection.
Enabling peopleEnablement
  • Partner activation programmeRecruiting workshops and multi-stage onboarding for AI products.
  • Onboarding in 15 minutesAn interview skill gives colleagues their own AI setup.
  • Shared internallyPresentation in German and English, setup guide, tool comparison.
Connected via MCP connectorsStack
Claude CoworkSalesforce · read-only, SOQLPardotNotionJiraGmailGoogle DriveWeb researchPython · sandboxn8nSubagents · parallel

Work samples

Eight workflows from the marketing team of a B2B software vendor that sells exclusively through partners. Customers, partners and internal figures are deliberately left out. Click a card to see problem, setup, result and rules.

Time figures for manual work are estimates. I'm happy to walk you through the setup live in an interview, using anonymised data.

Problem

Setup

Result

Rules
Skills

What I can do, and how you can tell.

Every skill with an honest self-assessment and a piece of evidence from practice. Not software engineering in the narrow sense: my strength is translating between the business side and technology.

Core

AI and automation

  • Agentic workflows 14 scheduled workflows in daily production.
  • Multi-agent orchestration Parallel research agents: 19 briefings in one day.
  • Context and prompt design More than 15 custom skills, style and profile files.
  • MCP and low-code Rebuilt an n8n flow: 882 false hits down to 2.
  • AI governance Human approval, sources required, read-only access.
Enabling people

Enablement and change

  • Partner enablement Activation programme with workshops and onboarding.
  • Training material Sales playbooks, one-pagers, self-assessment tools.
  • Presenting and explaining A core strength according to all three employment references.
  • Onboarding colleagues Onboarding skill with a 15-minute interview.
  • AI roadmap For a mid-sized company with around 80 employees.
Foundation

Go-to-market

  • Partner marketing Localised campaigns, events, partner acquisition.
  • Technical B2B content Articles, white papers, magazines since 2017, German and English.
  • Demand generation Lead campaigns, webinar and event funnels.
  • Positioning Digital sovereignty versus US platforms.
  • SEO and SEA Google Ads with budget responsibility for a web shop.
Toolbox

Data and systems

  • CRM and Salesforce SOQL queries, reports, data quality.
  • Reporting Daily dashboard from five CRM objects.
  • Knowledge management Notion knowledge base for people and AI.
  • ERP and master data Dynamics 365 BC, data flows between ERP, CRM and shop.
  • Python Reading and adapting scripts, no development.
Expert: used daily, built myself Advanced: regularly and independently Basics: still building
Tools Claude and CoworkMCPn8nSalesforcePardotNotionJiraGoogle WorkspaceHubSpotWordPressGoogle AdsDynamics 365 BC
Credentials

Proven, not claimed.

Excerpts from my German employment references (translated), plus my training since 2024. I'm happy to email you the full documents.

“The optimisations and AI-based approaches he introduced make an important contribution to the company's digitalisation and competitiveness.”
Interim reference, E-Commerce Manager · 2025 · translated
“He quickly gets to grips with complex systems and processes.”
Interim reference, E-Commerce Manager · 2025 · translated
“Thanks to his precise analytical skills and his remarkable ability to grasp things quickly, he was always able to assess even difficult situations accurately and immediately.”
Employment reference, Marketing Specialist · 2025 · translated
“The quality of Mr Rohmann's work was excellent at all times.”
Employment reference, Technical Editor · 2020 · translated
  • AI Fluency: Framework & FoundationsAnthropic
  • AI Automation with ClaudeAdvancing Women in Tech
  • AI competence for SMEsMittelstand-Digital (German Federal Ministry for Economic Affairs)
  • Where can AI fit into the business model?Mittelstand-Digital Centre Kaiserslautern
  • Generative AI fundamentalsMicrosoft and LinkedIn
  • Advancing AI in the organisationLinkedIn Learning
  • AI for content: prompting and imagesLinkedIn Learning, two courses
  • Inbound and Social MediaHubSpot Academy
Contact

Let's talk about your role.

A short message is all it takes. I'm happy to have a first video call at any time, and I'm available from mid-October 2026.

fabian.rohmann@gmx.net LinkedIn profile