Case studies

Deployments that moved the metric.

A closer look at how Nigerian and West African businesses ship AI and software with Reeksoft — the challenge, the approach, the outcomes.

AI-powered call centre with agents wearing green headsets at a bright Lagos workspace

24/7 voice AI for balance & transaction enquiries

The challenge

Peak-hour call queues at three contact centres routinely stretched past eight minutes. Over 60% of inbound calls were repetitive balance, mini-statement and last-transaction requests — draining agent time and inflating staffing cost across day and night shifts.

Our approach

  • Mapped the top 20 call intents against core banking APIs and IVR flows.
  • Designed a bilingual (English + Pidgin) voice persona with OTP-based caller authentication.
  • Integrated live to the core banking system for real-time balance, transactions and card controls.
  • Piloted with 10% of traffic before phased cut-over to full inbound volume.

The solution

A voice AI agent front-lines every inbound call, authenticates the caller via OTP, and resolves routine banking enquiries end-to-end. Complex issues — disputes, fraud, corporate banking — are warm-transferred to a human agent with the full transcript and caller context already on screen.

Outcomes

Lower staffing cost
55%
Hold time on routine calls
0s
Uninterrupted coverage
24/7
Caller CSAT on AI-only calls
92%
"Our night shift used to cost more than it earned. The AI agent quietly took over — and our customers barely noticed the switch."
— Head of Contact Centre Operations, Tier-2 Nigerian Bank
Smartphone displaying a WhatsApp AI assistant helping a customer track an order

WhatsApp AI for order tracking & support at scale

The challenge

The support inbox overflowed daily with repetitive order-status, delivery and returns questions from thousands of buyers. Response times had drifted past 12 hours, and negative reviews mentioning slow WhatsApp support were climbing week over week.

Our approach

  • Connected the WhatsApp Business API to the retailer's order management system.
  • Trained the agent on product catalogue, returns policy and delivery SLAs.
  • Added Pidgin and Yoruba conversational patterns alongside English for authentic tone.
  • Set clear handoff triggers — refunds, complaints, VIP customers — to route to humans instantly.

The solution

A WhatsApp AI agent reads live order data on demand, answers policy and product questions, processes return initiations and pushes proactive delivery updates. Human agents now handle only the 10% of conversations that genuinely need judgement.

Outcomes

Support tickets deflected
90%
First-response time
<1 min
To full rollout
30 days
Post-chat rating
4.8★
"Our team stopped drowning in 'where's my order?' messages. They finally have time to actually sell."
— Head of Customer Experience, Lagos E-Commerce Retailer
Glossy 3D emerald green chat bubbles floating on a soft light background

Web chat AI that qualifies loan & investment leads

The challenge

The site pulled strong traffic but converted poorly. Prospective borrowers and investors filled out interest forms then waited two to three days for an advisor callback — by which point most had gone quiet or signed with a competitor.

Our approach

  • Modelled qualification flows for each product line: personal loans, SME loans, investments.
  • Built product knowledge into the agent so it could answer eligibility and rate questions accurately.
  • Wired directly to the CRM to open a qualified lead the moment intent was confirmed.
  • Enabled calendar-based advisor booking inside the chat for the highest-intent visitors.

The solution

An embeddable chat widget greets every visitor, qualifies them against product rules, answers FAQs from the knowledge base, and either books a direct advisor call or hands off a hot lead to sales — all before the visitor leaves the page.

Outcomes

Higher lead-to-call
48%
Faster response
3×
FAQs auto-resolved
80%
Always-on qualification
24/7
"The AI qualifies leads better than half our junior team ever did — and it never sleeps."
— Chief Revenue Officer, West African Fintech
Developer workspace with a laptop showing code and mobile app mockups next to a green plant

Fleet & delivery platform with live tracking

The challenge

Dispatch ran on spreadsheets, WhatsApp groups and phone calls. SLA compliance had slipped to 71%, customers had no visibility into their shipments, and management had no reliable picture of fleet performance.

Our approach

  • Ran a two-week discovery with dispatchers, drivers and enterprise customers.
  • Shipped a driver mobile app and dispatcher web console in parallel two-week sprints.
  • Rolled out to one depot first, iterated for two weeks, then scaled to the full network.
  • Handed over full source code and a 30-day post-launch SLA.

The solution

A three-app platform: a driver mobile app for pickups, deliveries and proof-of-delivery; a dispatcher web console for routing and live monitoring; and a customer tracking page that removes the flood of 'where is my package?' calls.

Outcomes

SLA compliance
71% → 96%
MVP to production
10 wks
Client-owned code
100%
Fewer inbound tracking calls
40%
"We went from managing a fleet on WhatsApp to running it on real data. The delta shows up in every SLA report."
— Chief Operating Officer, Regional Logistics Operator

Ready to become our next case study?

Book a free discovery call and we'll map an AI or software rollout to your business — with the metrics you'd want to publish.