Colocation of call center services from On-Prem and eventual migration to Cloud across Kasha Rwanda and Kasha Kenya

In Detail

1) Business Problem with Context: When I joined Kasha Global, the call center was a single point of failure. Core Issues;

• On-prem call center infrastructure in Rwanda was running on E1 card technology, Hosted in an unreliable physical location, and was Dependent on unstable power and network connectivity.

• Frequent outages caused missed customer calls, Lost revenue, and severe customer dissatisfaction

• Kenya had better stability via Safaricom colocation, but Rwanda operations dragged overall CX down.

• Maintenance costs kept rising while reliability declined.

• The Freshworks CRM rollout exposed the need for modern, cloud-ready telephony.

Urgency: Customer support was mission-critical. Continued downtime directly undermined brand trust, regulatory expectations (health services), and growth plans.

2) Objective / Success Criteria

• Objectives;

  • Remove call center infrastructure from unreliable on-prem environments.

  • Colocate infrastructure in professional data centers.

  • Eliminate power and network-related outages.

  • Migrate telephony services to cloud-based infrastructure.

  • Integrate call center services with Freshworks CRM.

  • Reduce maintenance and operational costs.

• Success Criteria;

  • Stable call center operations across Rwanda and Kenya.

  • Near-zero unplanned downtime.

  • Seamless Freshcaller integration.

  • Reduced infrastructure and maintenance spend.

  • Scalable architecture for future markets (e.g. South Africa, and Democratic Republic of Congo).

3) My Role & Ownership: Program Owner & Technical Lead. I owned:

• Infrastructure assessment and future-state design.

• Vendor engagement with AOS (Rwanda), Safaricom (Kenya), and local In-country Telcos.

• Migration strategy (on-prem to colocation to cloud).

• SIP and telephony integration design.

• Stakeholder coordination (internal + external).

• Risk management and phased execution.

• Executive reporting and buy-in.

This was a high-risk, multi-vendor, multi-country program.

4) What I Did

• Phase 1 - Colocation

  • Audited existing on-prem telephony infrastructure.

  • Selected AOS data center in Rwanda for colocation.

  • Migrated call center servers from office premises to AOS.

  • Coordinated power, network, and redundancy setups.

  • Stabilized connectivity and eliminated location-based outages.

• Phase 2 - Cloud Migration

  • Initiated cloud telephony strategy aligned with Freshworks CRM.

  • Integrated SIP trunks with Freshcaller.

  • Coordinated MTN Rwanda cloud SIP rollout.

  • Worked with Safaricom Kenya for seamless cloud integration.

  • Decommissioned legacy hardware gradually.

  • Conducted testing and failover validation.

  • Trained support teams on the new system.

5) Technical Depth

• Architecture Evolution

  • Before: On-prem PBX (E1 cards) with which we had frequent outages

  • After: Cloud-based SIP telephony integrated with CRM

• Technologies

  • SIP trunking (MTN Rwanda, Safaricom Kenya)

  • Freshcaller (Freshworks Cloud Contact Center)

  • Data center colocation (AOS, Safaricom)

  • IP-based call routing

• Key Capabilities

  • Call logging and recording

  • CRM-linked call history

  • Scalable agent provisioning

  • Reduced hardware dependency

6) Results & Impact (Before vs After)

• Before;

  • Frequent, unpredictable downtime.

  • High maintenance costs.

  • Poor customer experience.

  • Operational firefighting.

  • No scalability.

• After;

  • Near-zero downtime over 12 months.

  • Stable, predictable call center operations.

  • Lower infrastructure and maintenance costs.

  • Seamless CRM integration.

  • Scalable cloud-first telephony model.

  • Improved customer satisfaction and agent productivity.

• Strategic Outcome: Customer support transformed from a liability into a reliable growth enabler.

7) Challenges & How I Solved Them

• Telco Delays (Rwanda): Extremely slow turnaround times. Solution was approached through securing executive sponsorship, maintaining constant follow-ups, and Structured phased milestones to show progress.

• Legacy Hardware Failures: Old servers continued to crash. Solution was approached via accelerated cloud migration, and minimized dependency on physical hardware.

• Multi-Vendor Complexity: Different vendors across countries. Solution here was Centralizing coordination, Single architecture blueprint, and Clear ownership per vendor.

8) What I’d Do Differently Next Time

• Secure buy-in earlier, Skip colocation, and move directly to cloud where feasible.

• Negotiate stronger SLAs upfront with Telcos.

• Budget earlier for cloud telephony.

• Pilot cloud services earlier to shorten timelines.

• Review and revisit exit clauses for existing underperforming vendors.