Six phases.
One accountable team.
The 912 engagement model in writing. The same delivery cadence our proposals use — published so you can size, scope, and validate against your own procurement process before the first phone call.
Why this model exists
Built by a founder who treats contracts as relationships.
Njuguna Waitara has run 912 since 2013 on a conviction most IT vendors talk about and few practise: an infrastructure contract is a relationship, not a transaction. The vendors who quote, install, invoice, and disappear are the reason African enterprises end up with five suppliers and zero accountability.
That conviction shows up in the numbers — the founder puts client retention at roughly nine in ten across 912’s history — and it shows up in this model. Every phase below exists because something breaks when it is skipped: proposals written without discovery, deployments without rollback plans, handovers without runbooks. The six phases are how a relationship survives contact with production infrastructure.
It is also why most clients consolidate with us over time — the One Contract model is the commercial shape of a relationship that already works.
Connect with Njuguna on LinkedInThe Universal Model
Every 912 engagement runs through these six phases.
The depth of each phase varies — a 2-day audit looks different from a multi-week SAP HANA DR — but the structure is identical. You always know where you are in the engagement and what comes next.
Discovery
Understand the environment before quoting.
We start with a structured conversation: what runs your business today, what hurts most, what compliance pressures you carry, and what you have already tried. We do not produce a proposal off a phone call. Every engagement begins with at least one site walk or read-only environment review so the scope reflects reality.
Deliverables
- Initial environment overview
- Stakeholder map and decision-rights
- Pain-point and constraint register
- Audit scope agreed in writing
Audit & Assessment
A 2-day deep dive that produces a written report, not a sales pitch.
On-site inspection plus remote scanning. We audit physical security (server rooms, environmentals, CCTV integration), networks (firewall rules, AD permissions, VLAN segmentation), cloud spend, backups (verified restores — not just scheduled jobs), and licensing posture. Every finding maps to a recommendation with priority and budget band.
Deliverables
- Day 1: discovery scans + physical inspection
- Day 2: policy, cloud, licensing review
- Executive findings briefing
- Prioritised, budgeted 3-year roadmap
Scope & Proposal
A written proposal you can hold us to.
Scope, deliverables, prerequisites, exclusions, fee schedule, validity period, payment terms, and acceptance criteria — all in one document. Most 912 proposals run 8–14 pages because we treat them as a delivery contract, not a brochure. Validity is 30 days from issue unless we are racing a hardware-price clock.
Deliverables
- Scope of work with explicit exclusions
- Fixed-price or transparent rate card
- Hardware and licensing line items separated
- Acceptance criteria and success metrics
Deployment
Phased execution with rollback at every step.
We do not do big-bang cutovers. Whether it is the Swing Migration protocol on virtualization, asynchronous HANA replication with a manual takeover gate, or weekend DNS cutovers on email migrations — every deployment is staged, validated, and reversible until the final go-live signal. Production stays online while we work.
Deliverables
- Phase-by-phase deployment plan
- Rollback procedure for each phase
- Pre-cutover validation checklist
- Communication plan for stakeholders
Testing & Runbook Handover
You receive documentation, not just access.
Every engagement ends with documented runbooks — failover procedures, restore tests, escalation contacts, configuration baselines, credentials in a managed vault. The same runbook your team uses on day two is the one our engineers use on day one. Six months of post-deployment support is included on most project engagements.
Deliverables
- Step-by-step runbook (failover, restore, escalation)
- Configuration documentation and architecture diagrams
- Credential handover via managed vault
- Admin training and shadow period
Managed Support
Continuous operation under a clear retainer.
Post-handover, most clients move into a Security Shield retainer (security monitoring) or full Managed IT (operational support). Retainer scope is defined in writing — what is included, what is project work, what is parts/hardware. Quarterly business reviews keep the roadmap honest as your environment evolves.
Deliverables
- Scoped retainer or AMC agreement
- Defined SLA and escalation path
- Quarterly business review cadence
- Roadmap updates as the environment changes
Per-Engagement Variants
How the six phases adapt per service.
The model is universal. The protocols within each phase differ by service — Swing Migration on virtualization, manual failover gate on SAP DR, weekend cutover on email migration. The highlights below come straight from the runbooks we hand over.
SAP HANA Disaster Recovery
Implementation in weeks; proposal valid 30 days
- Asynchronous HANA System Replication over MPLS
- Manual failover gate (split-brain prevention)
- ETIMS DR-site ready for tax-compliant invoicing
- Tested failover/failback runbook
Server Virtualization (Proxmox/KVM)
4–8 weeks end-to-end
- Three-phase Swing Migration protocol
- Production stays online throughout
- ZFS storage with bit-rot detection
- 6 months post-deployment support included
CCTV Surveillance
Multi-site assessments quoted per location
- Site visit and blind-spot mapping first
- Hardware procured against the assessment
- NVR commissioning + storage optimization
- Bi-annual maintenance recommended
3CX IP Telephony
Hosted PBX: 3 business days · Local Yeastar PBX: 2 days · Biometric door access: 1 day
- SIP and VLAN tagging configured
- Linkus for desktop/mobile extensions
- Call flow testing pre-cutover
- 3-month config support for 912 deployments
Power BI & Workflow Automation
First dashboard live in 2–4 weeks
- One focal question first, not a full platform
- Working dashboards every 2 weeks
- Underlying data engineering included
- 14% profitability lift in FMCG case study
Email & Workspace Migration
Background sync → weekend DNS cutover
- Pre-migration mailbox and DL audit
- SPF, DKIM, DMARC configured properly
- Final delta sync at cutover window
- Monday-morning go-live support
Founder’s Desk
Field notes from the engineer who signs the work.
Njuguna Waitara writes about the decisions behind 912’s engagements — why we build the way we build, and where most IT vendors get it wrong.
An infrastructure contract is a relationship, not a transaction. The vendors who quote, install, invoice, and vanish are exactly why so many African businesses end up with five suppliers and zero accountability.
Read on LinkedInWe design SAP HANA failover to be manual, not automatic — on purpose. An automated cutover that triggers on a network blip can split-brain your database and corrupt the very thing it was meant to protect. A documented runbook a human runs beats a clever script that fires at the wrong moment.
Read on LinkedInWe do not offer generic solutions. Every recommendation we put in front of a client is specific, justified, and deliverable — if we cannot defend why it is the right call for your situation, it does not go in the proposal.
Read on LinkedInWhat We Need From You
The client-input checklist — published, not buried.
Every 912 proposal includes a “Client inputs” block. We surface it on the public site too so you know what to gather before our first call — and so the audit doesn’t stall waiting on access requests.
For the audit
- Site access and stakeholder availability
- Asset inventory (current best understanding)
- Network diagrams or topology summary
- Admin credentials (read-only is fine for discovery)
- Current ISP, MPLS, and Wi-Fi coverage details
- List of incumbent vendors and contracts
For deployment
- Approved downtime windows and change windows
- Hardware procurement approval where applicable
- Security and application testing access
- Escalation contacts for incident scenarios
- Acceptance criteria sign-off
- For DR: latency benchmark figures from the MPLS provider
For managed support handover
- Owner for each in-scope system
- Reporting cadence preference (weekly, monthly)
- Preferred ticketing/escalation channel
- Compliance audit schedule (if applicable)
- Quarterly business-review attendees
Commercial Terms
The structure inside every 912 proposal.
No surprises in the contract. These are the commercial terms we use across SAP DR, virtualization, telephony, audit, and managed-IT engagements. Published so your procurement team can validate against standard expectations.
Quote validity
30 days from proposal date. Beyond 30 days we re-validate against hardware and licensing price movement before honouring.
Deposit
60% on Local Purchase Order (LPO) is our standard structure for project engagements. Balance on delivery and acceptance.
Hardware and licensing
Typically client-side per SAP, Fortinet, Microsoft, and Veeam commercial norms. 912 can broker on your behalf but the contract sits between you and the vendor.
Post-deployment support
Six months of post-deployment support included on most project engagements (virtualization, telephony, CCTV). Beyond that period, an AMC or retainer applies.
AMC parts and major service
AMC contracts cover scheduled inspections, firmware updates, and minor troubleshooting. Replacement parts and major service calls are quoted separately.
Liability cap
Where established, liability is capped at the subscription fees paid in the preceding six months. Standard SaaS commercial framing.
Published Delivery Timelines
How long each service actually takes.
Planning ranges drawn from delivered and scoped engagements, including hardware lead time, configuration work, dependencies, and any post-deployment support period. Your written scope confirms the actual sequence and duration.
| Engagement | Typical duration |
|---|---|
| Biometric door access installation | 1 day per door |
| Local Yeastar IP PBX | 2 business days |
| 3CX Hosted PBX setup | 3 business days |
| IT Audit (2-day deep dive) | 2 days on-site + 3–5 days for report |
| Email & Workspace Migration cutover | 1 weekend (background sync precedes) |
| First Power BI executive dashboard | 2–4 weeks from data access |
| CCTV system commissioning (single site) | 3–10 days depending on camera count |
| Server Virtualization (Swing Migration) | 4–8 weeks end-to-end |
| SAP HANA DR implementation | Multi-week, scoped against proposal |
Why This Process, Not A Generic One
The principles behind every 912 delivery choice.
The reason our proposals look unfamiliar isn’t the format — it’s the philosophy. Each principle below is a deliberate trade-off we’ve made against the obvious answer, because the obvious answer fails under African operating conditions.
Manual failover, not automatic
On SAP HANA DR we use asynchronous replication with a deliberate manual takeover gate. Automatic failover under flapping MPLS is where SAP environments lose data permanently — split-brain corruption is unrecoverable. We trade a few extra minutes of recovery time for a guarantee of data integrity.
Database-first, not CPU-first
Most ERP slowdowns are storage or memory bottlenecks, not CPU. Processor upgrades on bottlenecked storage don't move the needle. We profile the actual bottleneck before quoting hardware so you don't pay for the wrong fix.
VPN-first, not public ERP exposure
SAP and SQL servers are never reachable directly from the public internet on a 912 deployment. Administrators connect via VPN with AD authentication and 2FA, then route through a jump host with audit logging. Convenience never overrides isolation.
Tested backups, not scheduled ones
Backup jobs that run on schedule but have never been restored are how most ransomware incidents become catastrophic. Every 912 backup engagement includes restore tests in the first 30 days and quarterly thereafter.
Documented runbooks, not verbal handover
Every project ships with a runbook — failover procedure, restore steps, escalation tree, credential vault references. If the engineer who built it left for another job, your team can still operate the system tomorrow.
Honest residual-risk framing
We document what each engagement protects against — and what it does not. Asynchronous replication exposes seconds-to-minutes of data loss in normal link conditions. We say so up front rather than promising zero data loss we cannot deliver.
Phase 01 starts with a conversation.
Review the services before we map your environment, pain points, and the right next step in a 30-minute discovery call.