How Deep.bi reduced bench time by 34% in 8 weeks

NovaSoft US is a 62-person ITstaff augmentation firm based in Austin, Texas. They provide data engineers,cloud architects, and DevOps specialists to enterprise clients across the US onT&M contracts ranging from three to eighteen months. With a team that sizerunning 14 concurrent client engagements, resource visibility should have beena solved problem. It wasn't.

The delivery lead, Marcus, wasmanaging allocations in a shared Google Sheet updated twice a week — whenpeople remembered to update it. The real picture of who was available emergedslowly: through Slack messages, through end-of-week check-ins, through PMsflagging that someone had rolled off a project. By the time bench time wasvisible, it had already been sitting there for one, sometimes two weeks.

The financial impact wassignificant. A senior data engineer at NovaSoft costs the company approximately$140 per hour. Two engineers on the bench for a combined three weeks perquarter — a conservative estimate — represents over $50,000 in unrecovered costper quarter. Annualised, the problem was costing NovaSoft more than $180,000 incapacity that existed but was never converted into billable hours.

Marcus knew the problem. He didn'thave the tool to fix it. "We'd find out someone was free on a Friday andspend the weekend scrambling to find them something to bill against on Monday.Half the time we couldn't, and the week was gone." The Google Sheet toldhim what had happened. It couldn't tell him what was about to happen.

NovaSoft onboarded Synthelio inJanuary 2026 as a founding member. The first thing Marcus configured was theAllocation Gantt — importing their 62 employees, setting FTE percentages, andentering all 14 active project allocations. The setup took one day withSynthelio's onboarding team.

From day one, the Gantt showedwhat the spreadsheet never could: a 6-month rolling view of every employee'sallocation status, colour-coded by booking type. Solid teal bars for HardBooked. Dashed green for Soft Booked. And pale pink — the bench — clearlyvisible without any filtering or calculation. Marcus could see, on a Mondaymorning, exactly which engineers were rolling off in the next two, three, andfour weeks.

The billability percentage columnchanged the conversation. Instead of a weekly manual tally, NovaSoft'sleadership had a live number per person — updated automatically as allocationschanged. The threshold they set internally was 75%. Anyone below it appeared inamber. Anyone at 0% appeared in red. Marcus reviewed the view every morning.His weekly team meeting shifted from "who's available?" to"here's who we're proactively placing next."

They also activated the Demandtab. When NovaSoft's sales team had a qualified opportunity — a client likelyto close within 30 days — they logged it as a demand record in Synthelio: role,seniority, technology, and required start date. Delivery could see the pipelinedemand before the deal closed. In two instances in the first 8 weeks, they hadthe right engineer pre-identified and introduced to the client within 24 hoursof the deal closing.

After 8 weeks on Synthelio,NovaSoft's average team billability had risen from 58% to 78% — a 20 percentagepoint improvement. Bench time, measured as total unallocated hours across theteam per week, fell by 34%.

The change came not from hiringdifferently or winning more clients — their pipeline was consistent throughoutthe period. It came from seeing the bench form earlier and acting faster.Marcus estimates that the average lag between an engineer rolling off a projectand being reallocated to a new one fell from 11 days to 4 days.

The Demand tab produced anadditional benefit NovaSoft hadn't anticipated: sales started closing dealsfaster. When a client asked "can you have a senior Azure architectavailable by March 15?", sales could answer in the same call — because theycould see the allocation view themselves. Two proposals that had previouslystalled on resourcing questions were closed within a week of Synthelio goinglive.

•    Bench time reduced by 34%in the first 8 weeks

•    Team billability average:58% → 78%

•    Average reallocation lag:11 days → 4 days

•    $180K+ in estimated annualcapacity recovered

•    Two sales proposalsaccelerated by real-time resource visibility