From spreadsheets to AI-built workflows — in four weeks
How Paessler's finance team left a bootcamp with working Claude-based workflows of their own — not a slide deck.
Across GL/Accounting, FP&A and Procurement — Warsaw + US
On the specific tasks participants automated, first re-run
Working Claude-based workflows built during the programme
Want to see what we'd build for you?
The Situation
Paessler AG is a network and IT infrastructure monitoring software company headquartered in Nuremberg, Germany, best known for its PRTG platform. Its finance team — GL/Accounting, FP&A and Procurement — spans Warsaw and the US, and runs the workflows every finance team knows: trial balance consolidation, accruals, expense refreshes, contract review, pipeline-to-forecast. Before the programme, most of that ran in Excel, by hand, every month. Some team members were already daily Claude users; several had never opened it. Kris Pakula, SVP Finance, wanted the team to move past ad hoc AI use into something structured and safe — owned by the people doing the work, not a tool bought once and left on a shelf.
The Challenge
- Core monthly workflows — trial balance consolidation, accruals, expense refreshes, contract review, pipeline-to-forecast — largely run by hand in Excel
- Uneven AI adoption across the team: some already daily Claude users, others had never opened it
- Team split across Warsaw and the US, needing sessions and support that worked across time zones
- Leadership wanted structured, safe AI use owned by the people doing the work — not a one-off tool rollout
Our Approach
Four weeks, built on Claude throughout. incro ran three live 90-minute sessions plus a half-day build day, using incro's CRAFT prompting framework (Context, Request, Artefact, Filters, Tripwires) to teach structure, not shortcuts. Every session ended with participants building against their own real workflows:
- Week 1 — Tool orientation and an AI-readiness matrix, scored live against Paessler's actual workflows.
- Week 2 — A hands-on Skills-building workshop, plus the responsible-use rules Paessler's finance team follows today.
- Week 3 — On-site build day: GL and FP&A sub-teams, real Paessler data, one working Claude Project per person.
- Week 4 — Live demos, a dedicated Teams channel, and best practices share.
What We Built
- FP&A forecast prep — Pipeline-to-forecast formatting, cut from about 30 minutes to about 10.
- GL vendor/compliance checks — VAT ID checks, cut from about 30 minutes to about 10, and a route into daily Claude use for someone who'd never used it before the programme.
- Reconciliation & cash reporting — GL export analysis and weekly cash-balance consolidation across entities, cut from 2–4 hours to roughly 1–2.
- Monthly expenses refresh — A multi-part Monthly Expenses Refresh, rebuilt as a Claude Project with reusable Skills for the recurring pieces.
- Flux / variance analysis — An actual-vs-actual flux analysis flagging variances at account, vendor and department level, demoed live in Week 4, built for the July close.
Conclusion
Four weeks in, Paessler's finance team isn't just aware of what Claude can do — they own working workflows built against their own data, in production, without needing incro in the room. The next step is a +30-day follow-up to size the impact against a full monthly cycle, and a shared Teams channel to keep the practice compounding.
Your financial data won't fix itself.
30 minutes. We'll tell you exactly where your data is costing you money — and what AI can do about it.

.png)