-May-15-2026-08-50-57-5826-PM.png?width=127&height=45&name=image%20(4)-May-15-2026-08-50-57-5826-PM.png){kind=small}
The Background
Accelerating Quality of Service using Technology
Bitfocus is a software company in a mission-critical corner of public service. Its flagship product, Clarity Human Services, is a Homeless Management Information System (HMIS) used by seven of the ten U.S. cities with the largest homeless populations, plus hundreds of Continuums of Care, counties, and nonprofits coordinating shelter, outreach, and housing services across the country.
What sets Bitfocus apart in a small, tightly connected industry isn't price or marketing reach. It's the quality of service the company delivers to the communities it administers. Federal funding cycles, HUD reporting deadlines, jurisdiction-specific policies: Bitfocus handles the technology so frontline agencies can focus on the people in front of them.
That's why the company's reputation rides less on its product roadmap and more on the experience its customers have when something breaks. And as Bitfocus's customer portfolio has grown, the demands on that experience have grown with it.
The Challenge
When a ticket waits, a service stalls
The end users behind a Bitfocus support request are rarely sitting at a desk. They're street outreach workers logging an encounter from a tablet, shelter staff trying to enroll a family arriving at intake, program administrators racing to close out a federal grant report. Homelessness is a 24-hour problem, and the systems that respond to it can't pause for time zones or business hours.
When Clarity slowed down, or when a question went unanswered until the next business day, the impact didn't stop at the help desk. It rippled outward to the people those frontline workers were trying to serve.
A West Coast desk serving a continent
Inside Bitfocus, that pressure landed on Jeffrey Ugai, Chief Operating Officer, who oversees the service function. His team is West Coast based but supports a customer portfolio that stretches from the Bay Area to the Eastern Seaboard. East Coast mornings began with a single agent on duty.
Weekend tickets could sit until Monday. Roughly 75 to 80 percent of inbound tickets came from end users — frontline workers rather than the system administrators those workers reported to — which meant volume was high, varied, and unpredictable.
The team needed to close the time-zone and after-hours gap. But not by hiring around the clock, because Bitfocus's customer portfolio was expanding faster than its support headcount could reasonably keep up. And not at the cost of accuracy, either.
Every Continuum of Care has its own policies, scoring criteria, and reporting cadence; a generic answer to a Los Angeles user could be the wrong answer for a user in Cuyahoga County.
What Jeff needed was coverage that was fast, accurate, and community-aware. What he had was a queue.
The Solution
Putting an always-on teammate in front of the frontline
The fix had to start where the customer impact was: with the outreach workers, shelter staff, and administrators who needed answers in the moment, regardless of what time it was on the West Coast.
Bitfocus deployed HubSpot's Breeze Customer Agent, internally nicknamed "Rambo" after the founder's dog, across two channels where customers actually showed up: the company's HubSpot-hosted help center and the live chat embedded inside the Clarity Human Services application itself. Rambo went live 24/7, drawing on the same knowledge base the human team relied on, and it began handling routine, templatable questions the moment they arrived. Password resets at 4 a.m. Eastern. Reporting questions on a Sunday. Configuration questions during a shelter intake rush.
Because Bitfocus was already on Service Hub — with its help center, ticket pipeline across email, phone, and chat, and Help Desk workspace all running in HubSpot — Customer Agent wasn't a separate tool to procure or stitch together. It was a natural extension of the system the team already knew. "Having everything in one system made HubSpot the obvious choice over a standalone AI tool," Jeff explained.
A diagnostic tool, not just a deflection tool
Rambo's value, the team quickly discovered, wasn't only in the tickets it resolved. It was in the ones it stumbled on. Every time the agent struggled with a question, it surfaced a gap in Bitfocus's documentation: a missing article, content drift, an under-explained workflow. The team had spent years trying to identify those gaps manually, with limited visibility. Now Rambo was pointing right at them.
The knowledge base and the agent began improving each other in a tight loop. As documentation got sharper, Rambo got more accurate. As Rambo got more accurate, the human team had cleaner content to pull from in their own replies — something Jeff's team had previously tried to encourage manually with little traction.
Community context where the work happens
For tickets that did need a human, Bitfocus paired the Help Desk workspace with custom objects (for community instances, accounts, and homeless response system attributes) so agents could see jurisdiction-specific context inline as they worked a case. The same ticket about a coordinated entry assessment now arrived with the right Continuum of Care policies one click away. Jeff has called this combination "crucial to the success overall of Help Desk."
And for the tickets that sat in the middle — not fully deflectable but repeatable — the team adopted reply recommendations, where Customer Agent drafts a response and a human agent edits and sends. Drafts come out cleaner. Replies go out faster. Follow-up tickets drop. It's the quieter productivity story sitting alongside the headline deflection number.
The Transformation
Frontline workers get answers when they need them
The clearest signal that something had changed wasn't in a dashboard. It was in customer behavior. Shelter staff and outreach workers who used to call Bitfocus by default started reaching for chat first. Earlier attempts to nudge customers in that direction hadn't taken hold.
What changed was that chat now gave them a meaningful answer in the moment, day or night. The shift was real enough that customers who had initially been skeptical of an AI agent — and in an industry where everyone knows everyone, skepticism travels — became advocates as Rambo's accuracy improved.
That ripple ran straight back to the people Clarity was built for. Fewer overnight blockers for outreach teams. Fewer Monday-morning backlogs for shelter intake. Fewer moments where a frontline worker had to tell someone in front of them, "the system's down, come back later."
The numbers behind the shift
Inside Bitfocus, the impact has been just as concrete. Rambo now deflects 40% of incoming tickets, with a 65% reduction in overall time-to-solve and roughly $0.3M in savings. Single-touch resolutions are up. Reopens are down.
And in what Jeff considers the most telling number, ticket volume has plateaued even as the customer portfolio has continued to grow, meaning each new community Bitfocus brings on no longer adds a proportional load to the support queue.
Just as importantly, the team that lived through the shift didn't lose ground in the process. Service team eNPS is among the highest in the company. The agents kept the work that needed their judgment — the nuanced, community-specific tickets only a human can handle — and offloaded the work that didn't.
A backbone for what's next
With Service Hub, Customer Agent, the help desk, and custom objects all running in one place, Bitfocus is now extending the same instinct outward. Engineering and product are piloting AI to mine ticket data for product feedback.
The website has been rebuilt on Content Hub, with IP-based routing tying customer signals back to the CRM. The integration thesis Jeff bet on early is paying off in places beyond support.
For Jeff, the deeper change is in how the team thinks about its own work. Every gap is a content gap to close. Every stumble is a chance to make the next answer better. The support function he runs is now absorbing growth instead of being squeezed by it.
-May-15-2026-08-50-57-5826-PM.png?width=135&height=48&name=image%20(4)-May-15-2026-08-50-57-5826-PM.png){kind=small}
-Apr-13-2026-05-55-20-9731-PM.png?width=215&height=50&name=image%20(1)-Apr-13-2026-05-55-20-9731-PM.png)
.png?width=215&height=50&name=RevenueWell_Logo%20(1).png)
