Tax teams are expected to manage increasing regulatory complexity, growing data volumes, expanding global obligations, and tighter deadlines — all while operating with lean teams and fragmented systems.
The volume of work compounds every cycle. The tooling was built to organize it, not to carry it.
Traditional software helps teams manage work. It doesn’t execute it.
Introducing Derive
A new operating model for enterprise tax.
Derive combines AI agents, tax specialists, and your team to execute tax work across the entire tax lifecycle — from recurring compliance to the most complex international workflows.
Instead of giving teams another tool, Derive becomes an operational layer that helps work get done.
Inside a single obligation
Work executed end‑to‑end — with your team in command.
01Learns02Runs03Escalates04Approves05Retains
Meridian Global US, Inc.
FORM 5471 · SCHEDULE J · ROW 07 · §986
Learning
SourcesTrial balance · Prior filings · GL feed
Ownership100% CFC · US parent
TreatmentSubpart F · §951A GILTI
Derive ingests the entity’s data and prior filings, and learns the treatment.
SCHEDULE J — EARNINGS & PROFITS · USD
Opening balance4,812,000
Current-year E&P1,190,400
Row 07 · §986 FX remeasurement(214,700)
Closing balance5,787,700
An AI agent builds the workpapers and drafts the return — 14/14, conf 96.4%.
ExceptionRow 07 · §986 FX remeasurement
ROR. Okafor · International Tax
“Spot vs. weighted-average rate diverged over 5% this period. Recommend the weighted-average method per policy TP‑04; net impact (214,700).”
Row 07 is routed to a specialist — reasoning attached, awaiting your call.
A new operating model for enterprise tax teams. Derive combines AI agents, tax specialists, and your existing team into a single managed-operations model. AI agents execute work, specialists provide domain expertise and oversight, and your team retains control of critical decisions and approvals.
“Spot vs. weighted-average rate diverged over 5% this period. Recommend the weighted-average method per policy TP‑04; net impact (214,700).”