Durban Lending Rollout

Evening demand from Durban applicants changed the brief

The client was a Durban lending business handling applications from borrowers across KwaZulu-Natal and other metro areas. Their main pressure point was the evening rush. People completed online applications after work, then reached the identity step when internal staff were already stretched. They wanted a remote video verification process that felt orderly, kept the evidence trail intact and did not slow account opening to a crawl.

The brief was practical. Staff needed a way to run guided video sessions, capture the ID document clearly, compare the face on screen with the document photo, and note the outcome in a consistent format. The lending team also wanted clear escalation rules for blurred images, mismatched details, dropped calls and cases where a customer could not complete the liveness check cleanly.

• Client type: online lending business
• Town: Durban
• Demand pattern: after-hours applicant traffic
• Core need: remote identity verification with recorded evidence

Working around short windows, mixed devices and weak connections

We had to shape the rollout around several constraints. Applicants were joining from different phones and laptop cameras, some with limited space and uneven lighting. The team also needed a process that could be taught to different agents without the script drifting from one person to the next. Record keeping had to stay consistent, because the lending file needed a clear trail for consent, document handling and the verification outcome.

We kept the moving parts tight. Session instructions were shortened, the ID capture sequence was written in plain steps and the checks were ordered so the agent could keep control of the call without sounding mechanical. We also added handling notes for retakes, failed face matches and incomplete sessions, which kept the evening queue from becoming messy.

• Mixed device use from mobile and benchtop applicants
• Variable lighting in homes and offices
• Need for consistent agent wording
• Clear records for each completed session

What we put in place for the Durban rollout

We started with workflow mapping, then moved into integration planning with the client’s application stack. From there we aligned the identity step to the lending journey so that the applicant would meet the video check at the right point, not too early and not after too many fields had already been submitted. API coordination covered the handoff between the application record and the verification file, while our script design team shaped the agent prompts for ID display, face comparison and consent capture.

Training sessions covered the practical side of the job. Agents were shown how to slow the pace without losing control, how to ask for a fresh camera angle and how to write notes that another team member could read later without guessing. We also helped the client build an escalation design for difficult sessions, plus a quality monitoring plan that made spot checks easier across the evening shift.

• Workflow Mapping
• Integration Planning
• API Coordination
• Agent Script Design
• Staff Training
• Escalation Design
• Quality Monitoring

Six weeks from mapping to live evening cover

The rollout ran over six weeks. The first stage was discovery and process mapping, where we traced the applicant path and marked the points where a video step made sense. The second stage was script and integration work, then a pilot launch with a small group of agents handling live cases under supervision. After that, we adjusted the notes for edge cases and set the monitoring pattern for the full evening shift.

That sequence mattered because the team did not need a heavy redesign. They needed a clean working line that could carry real traffic without confusing applicants or staff. The pilot showed where applicants paused, where agents needed stronger prompts and where the handover from application to verification file needed tidier wording.

What changed once the evening queue had structure

Once the process settled, the lending team had a steadier way to handle after-hours applicants. Agents could run the session with the same steps each time, which reduced drift in how identity checks were recorded. The client also gained better visibility over failed sessions, incomplete uploads and cases that needed a manual follow-up before the account could move forward.

The biggest shift was operational rather than cosmetic. The team could manage more evening checks without improvising each one, and the file for each applicant was easier to review later. That helped with internal oversight and made staff training simpler too, because new agents were learning one process instead of four different versions of it.

• More even handling of after-hours demand
• Cleaner session notes for review and audit trails
• Less agent drift in script delivery
• Clearer escalation of mismatches and retakes

What we would use again in a similar lending rollout

The strongest lesson from this job was to design around the real hour when the queue gets busy, not around office hours on paper. A lending team working with online applicants needs a verification flow that agents can run under pressure without sounding rushed or vague. We would use the same approach again for a business that needs guided video sessions, document capture, face checks and a reliable record of each completed step.

We would also keep the training close to the live workflow. That meant building the script, the escalation path and the monitoring notes together, rather than treating them as separate pieces. When those three parts match, the evening team can carry the workload with less friction and fewer handover errors.