A two-day AI workshop that gave LANXESS India's finance shared services hands-on tools for close, tax, treasury and payables.
Two-day finance AI workshop for 25 LANXESS India professionals across Record-to-Report, Accounts Payable, Accounts Receivable, Treasury, Direct and Indirect Tax and Controlling.
The starting point.
LANXESS India's finance team runs accounting, payables, receivables, master data and tax for multiple legal entities across India, APAC and EMEA. Before the workshop, participants described their pain points in a pre-programme survey: customs-duty documents checked by hand before posting in S/4HANA, bank-portal uploads for import payments, TDS and 26AS reconciliations, state-wise P&L for GST, and MIS dashboards rebuilt from many Excel files every period. The team already ran RPA bots, so the brief was to add generative AI on top of existing automation, not to start from zero.
How the engagement was structured.
- Used the pre-programme survey to map each participant's role (R2R, AP, AR, Treasury, Tax, Controlling) and their stated problem before designing the labs.
- Day 1: role-specific AI personalisation, custom research assistants, Schedule III financial statements and a financial dashboard generated from a trial balance, NotebookLM on the LANXESS annual and quarterly reports, and offline tools built with AI, including an import-documents tool and an MIS dashboard.
- Day 2: Excel VBA automation for file indexing and data reconciliation, Outlook mail automation, a Copilot agent built on the LANXESS India chart of accounts, AI anomaly detection on sales-invoice data, and AI-generated presentations in the LANXESS brand template.
- Closed with AI risk, data protection and regulation so new tools fit the team's ICFR and audit requirements.
What changed.
- Participants built working tools during the labs, including a trial-balance-to-Schedule III generator, a chart-of-accounts assistant and an invoice anomaly checker.
- Tax and treasury teams left with prompt workflows for reconciliations, notice research and document reading that previously needed manual effort.
- The team gained a shared AI approach that sits alongside its existing RPA bots, with clear rules on data handling.