Building Navi's Reconciliation Engine
Chose a third data source to improve reconciliation before expanding manual tooling.
- Reconciliation
- Payments infrastructure
- Operations tooling
- Data-backed prioritization
Astha Gupta
Senior Product Manager at Zeta
From their time as
Associate Product Manager
Navi • 2022 - 2023
Overview
Astha helped build Navi's reconciliation engine and chose a third data source over immediate manual-resolution tooling. The resulting engine reached 99.12% accuracy.
The Story
Astha worked on payments at Navi while the team was building a reconciliation engine alongside requests from finance and operations for manual-resolution tooling. The product decision was whether to build the operations portal first or improve automated matching.
She evaluated unreconciled cases against a third data source. The analysis showed how many additional items could be reconciled through integration, creating a clearer long-term path than starting with manual intervention.
Astha used that evidence to align finance and operations on sequencing. The manual portal could follow later, while the team integrated the third source to improve the engine's ability to reconcile exceptions.
The reconciliation engine reached 99.12% accuracy, reducing the need for manual reconciliation at scale.
Ownership Snapshot
Broad role
Associate Product Manager across payments infrastructure.
Goal
Improve reconciliation quality and reduce manual operational work.
Direct ownership
Built the case for third-source integration and aligned stakeholders on sequencing.
Team execution
Worked with finance, operations, and engineering on the reconciliation solution.