Natural-language access to procurement and revenue-contract execution data, plus outside-counsel spend by firm and practice area. The first time leadership could ask a question and get an answer without pulling data across four different systems.
Legal leadership needed to answer basic operational questions: how many NDAs did we execute last quarter? What are we spending with each outside counsel firm? Which practice areas drive the most contract volume?
Answering any of these required manual pulls across four disconnected systems — Tableau for contract data, SFDC for revenue contracts, SAP for procurement, SimpleLegal for outside-counsel spend. Each query meant logging into a different tool, running a different report, and manually reconciling the results. A single question could consume hours of analyst time. And the answer was stale the moment it was produced.
A Glean agent connected to all four source systems. Leadership asks questions in natural language — "what did we spend with Firm X last quarter across all practice areas?" — and the agent queries the relevant sources, reconciles the data, and returns a structured answer with line-item detail.
Key data domains: procurement contracts (MSSA, NDA, sales orders), revenue contracts (by customer, region, type), outside-counsel spend (by firm, practice area, matter type). The agent knows which source to query based on the question — the user doesn't need to know where the data lives.
Improvement: Hours of manual data pulls and reconciliation across 4 systems → seconds-long natural-language query. The analyst role shifted from report runner to data steward. Leadership got answers while the question was still relevant. The natural-language interface meant non-technical stakeholders could self-serve for the first time.
ROI: ~190 analyst-hours/year at a ~$100/hr loaded analyst cost — roughly $19K recovered annually.