THE DISPATCH
ORIGINS A TAXI TEAM BECOMES A BUSINESSNOW AI + HUMAN INTELLIGENCEON THE LINE TRANSPORTATION / ECOMMERCE / SAAS

Company / The outsourcing business

Pac Biz kept the calls. And built a business.

A taxi company’s phone team became Pac Biz Outsourcing. Now the business is applying AI to the work behind the headset, while keeping people responsible for the conversation.

Advertisement

The curious thing about Pac Biz Outsourcing is that its founders originally had cars to worry about. Their Scottsdale taxi business needed people to answer calls. A team in Dumaguete, Philippines, did that work. Then the supporting operation acquired a life of its own: other transportation companies needed the same thing. In 2015, Eric Mulvin and Bob Grunstein made the call center the business.

The story in three stops
  • A taxi operation produced a service other fleets could buy.
  • Pac Biz sells agents with training, supervision and quality control.
  • Its AI work targets the time spent reviewing and improving calls.

That origin is useful because outsourcing can sound terribly abstract. A rider needs a booking. Someone must understand the request, use the dispatch system and get the details right. The phone conversation is part of delivering the ride. Pac Biz learned customer support in a business where the handoff had wheels.

01 The department that escaped

For its first year, the new call-center operation had five employees. By October 2018, Pac Biz reported more than 145 staff. Its customers extended across the United States, Canada and the United Kingdom, with Australia next in its plans. The company’s story includes call takers moving into operations management and quality assurance. Knowledge accumulated on the phones became something the company could teach.

Pac Biz staff wearing headsets at workstations in a historical company photograph
Small headset. Considerable responsibility. A historical Pac Biz staff photograph. The booking only looks simple from the passenger’s end.

Its transportation specialization remains a concrete distinction in a crowded outsourcing market. Pac Biz describes itself as an approved iCabbi contact center, with agents familiar with the platform’s booking and dispatch work. A fleet buying that service is purchasing familiarity with an operating environment, alongside conversational English and phone coverage.

The wider menu includes customer service, email and chat support, virtual assistants and back-office work. Today’s website addresses transportation, SaaS and ecommerce businesses. Its homepage includes a Smashtech testimonial. The common customer is a business whose growing queue of requests has become a staffing and management problem.

02 What sits behind the hourly rate

Consider a financial-services customer whose review appeared in November 2017. The firm handled about 50 inbound calls a day. It wanted its sales representatives making outbound calls, so Pac Biz took the incoming ones. The customer reported saving 20 hours a week and $2,000 a month. It also offered a pleasingly practical criticism: “Their call clarity could improve.”

One customer’s account / 2017
20hours / weekStaff time saved
$2kper monthReported savings

Financial-services reviewer. First-year spending: $10,000-$49,999. Historical, self-reported results; individual outcomes vary.

The customer paid hourly. Pac Biz’s company profile describes a flat monthly fee per agent, paid in the client’s currency. Those are service purchases, with staffing and management attached. Comparing them requires counting the work a local hire creates: recruiting, training, scheduling, supervision and cover when somebody is absent.

The useful comparison is between complete arrangements. An in-house team gives a business direct control and local knowledge. A managed provider brings recruiting and supervision into the purchase. Automated voice systems take another slice of the work. Each still needs rules for unusual requests. The buyer should compare coverage, escalation and management effort alongside the invoice.

A service document on Pac Biz’s website spells out that surrounding machinery: a dedicated 40-hour agent, account management, supervision, quality assurance, training and company-culture support. It describes power and internet redundancy, plus check-ins after 30 and 60 days. The package also offered two to three weeks of free agent training. These details reveal what the company thinks clients are actually buying: continuity.

03 Listening became a software problem

By 2025, Mulvin was describing another problem: growth had plateaued despite marketing efforts. He said the company paused its marketing to reconsider its direction as AI changed the market. That admission makes the subsequent product work more interesting. An established support provider had to examine which parts of its own operation software could change.

Pac Biz’s answer includes a QA application. The company describes automated call transcription and scoring, performance dashboards, coaching insights, role-based access and a dispute process. That last detail deserves attention. An agent needs a way to challenge a score if automated judgment misses context. Quality control itself needs quality control.

There is also a marketing-call scoring tool using red, yellow and green signals for lead quality, and transportation tools in development. At Mobilize 2025, Pac Biz discussed software for scoring taxi and NEMT calls. The operational logic is straightforward: reduce time spent reviewing recordings so managers can spend more time helping agents.

“Their call clarity could improve.”

A financial-services customer, 2017

04 Keep someone responsible

A Pac Biz staff member wearing a headset and company shirt
The human end of the line. Staff portrait published by Pac Biz.

Pac Biz’s own AI guidance is unusually useful here. It stresses clean data, testing models and human supervision. Its virtual-assistant offering similarly puts process documentation alongside task completion. The assistant is expected to learn the client’s systems and help make recurring work repeatable.

Its published values include compassion, family, respect, teamwork and personal growth. Great Place to Work’s company profile reports 89% positive employee sentiment against a 65% Philippine benchmark. That supports the culture story, although it does not measure the quality of a particular customer’s calls.

For buyers, the copyable lesson is to define the handoff before adding capacity. Write down the request, the decision an agent can make and the point at which somebody else must intervene. Choose a measurable result, then review actual calls together. That is a practical reading of Pac Biz’s emphasis on documented workflows and calibration.

A sensible starting assignment would have a bounded queue and an explicit owner: incoming booking requests, order questions or appointment scheduling. That makes mistakes visible and coaching specific. Throwing an undocumented mixture of urgent tasks over the wall makes the service harder to judge. Pac Biz’s emphasis on learning a client’s tools matters most when the client can explain how those tools should be used.

The arrangement depends on compatible systems, clear instructions and someone owning exceptions. Pac Biz itself warns about integration problems and unreliable AI output. Its October 2026 publishing turns to protecting NEMT rider information. Sensitive data and difficult conversations keep responsibility firmly in the picture. A business can delegate a ringing phone; it still has to decide what a good answer sounds like.