A hundred and twenty thousand accounts before the deadline.

The situation

The firm recovers utility costs for landlords and property managers through municipal refund and credit programs. Eligibility was checked by hand, one account at a time, on municipal websites, against city deadlines that don’t move.

What we found

One process was hiding two problems. Existing client accounts weren’t re-screened often enough, so savings that appeared when programs or eligibility rules changed were missed, because nobody had the hours. And there was no way at all to find new clients, so growth depended entirely on referrals.

The same lookup, run at scale, solves both. You notice that by looking at the business as well as the task.

What we built

• Scripts that run the municipal lookups at volume, replacing account-by-account checking.
• Two ways out: results fill a searchable web dashboard, or merge with the client’s own database and come back as spreadsheets their systems already read.
• Room to grow: more agent machines start up when a deadline needs a faster turnaround.

The result

About twenty accounts a minute per agent machine, and about eighty a minute across four agents at peak. Runs cover 30,000 to 120,000 accounts depending on the program, and they finish inside the municipal deadline.

For scale: at one lookup a minute by hand, a 120,000-account run is about 2,000 work-hours, roughly a year of one person’s time. Work at this scale was never possible by hand.

The engagement is ongoing. Operating Lift maintains and runs the scripts, and capacity grows to whatever a deadline requires.