Lifetime Citizen Portal Access — AI Briefings, Alerts & Unlimited Follows
Finance staff demo FloQast AI reconciliation tool; council asks for city AI policy before contract
Loading...
Summary
Pocatello finance staff demonstrated FloQast, an AI‑assisted reconciliation tool, to the council. Staff said the product could cut reconciliation hours significantly; council requested an AI use policy before any procurement.
The Pocatello finance department invited FloQast representatives to demonstrate AI‑assisted reconciliation software and to discuss its potential to streamline the city’s month‑end and bank reconciliations.
Theron Hill of the finance department framed the need: daily cash reconciliations for a single operating bank account must be mapped across many funds, a manual process that he said consumed 70–80 hours per month for a single staffer. FloQast representatives demonstrated an automated matching engine that uses trained rules to group line items, match batches and recommend deterministic rules that can be edited by accountants.
Albert Loveri, senior solutions consultant for FloQast, and Alec Perizzo, an account executive, joined the presentation. FloQast staff said the product integrates with the city’s Central Square ERP and that the company enforces encryption and data‑segregation practices; FloQast representatives noted it maintains SOC reports and an AI‑specific certification (ISO 42001) and said customer data are not used to train external models. In the demonstration the tool produced an initial match rate of around 56% on a representative dataset; FloQast and staff said typical long‑term match rates can approach 90% as rules are refined and after planned ERP reporting changes are in place (an upgrade the city expects in October).
FloQast outlined a typical implementation timeline of roughly 3–6 weeks, requiring an estimated 6–12 hours of one accounting resource and 2–4 hours of IT time for setup. The company said implementation fees are one time and that implementation support is included for the life of the contract; the vendor said historical data and matched rules are exportable if the city later discontinues the service. Councilors asked about error rates, “hallucination” risk and handling exceptions; FloQast representatives described a conservative, reviewer‑based workflow and said the system produces human‑reviewable matches so false positives are minimized.
Council members praised the potential efficiency gains but asked staff to prepare an AI policy before presenting any contract. Theron Hill said staff are drafting an AI policy and that any procurement would return to council after legal and policy review. Finance staff did not request immediate council action at the work session.
