Fixed scope project
A defined outcome with a defined end. Suits an audit, a single migration or a recovery rebuild where the boundary is clear.
- Priced against the stage plan
- Milestone based
- Handover with runbooks
How We Work
Every stage has to survive the one that follows it.
Four stages, each signed off before the next begins, and each producing something the business can act on by itself. Funding that arrives in tranches should never strand the work halfway.
Operating rules
Three rules that hold across every engagement
Rule 01
Inventory precedes design
Nothing is drawn until we know what is actually running, including the undocumented parts. A design built on an inaccurate picture is a plan to find problems during cutover.
Rule 02
Rollback before schedule
Every cutover has a tested path backwards before it is given a date. If the reverse has not been rehearsed, the forward move does not get committed.
Rule 03
Each phase stands alone
A stage that only pays off if the next one is funded is a hostage, not a plan. Every phase leaves the estate defensibly better than it found it.
The engagement
Four stages, inputs to outputs
Stage 01
Assess the estate
A full inventory of what is running, mapped to what depends on it. End of support exposure is ranked by business impact rather than by asset age, so the list you get back is ordered by what actually falls over first.
- Inputs
- Existing asset registers, however stale
- Network and rack access
- Change and incident history
- Activities
- Physical and logical discovery
- Dependency mapping
- Vendor lifecycle lookup
- Power and cooling baseline
- Outputs
- Risk register, ranked
- Verified estate inventory
- Remediation sequence
Typical duration3 to 5 weeks
Stage 02
Design the target
Target architecture across cloud, storage, recovery and power, costed and sequenced. Energy is priced into the design rather than assessed separately, because it is now large enough to change which option wins.
- Inputs
- Stage 01 risk register
- Business continuity requirements
- Budget cycle and constraints
- Activities
- Workload placement modelling
- Recovery objective setting
- Platform selection
- Phasing and cost modelling
- Outputs
- Target architecture
- Phased migration plan
- Tested rollback positions
Typical duration4 to 6 weeks
Stage 03
Deploy in stages
Work lands in increments the operation can absorb. Recovery times are proven under load before anything carries production traffic, and every window has a rehearsed way back.
- Inputs
- Approved migration plan
- Change windows
- Rollback criteria
- Activities
- Build and integration
- Restore rehearsal under load
- Staged cutover
- Hypercare
- Outputs
- Proven cutover, per phase
- Evidenced recovery times
- Updated runbooks
Typical duration3 to 9 months
Stage 04
Operate and advise
Managed support, scheduled recovery rehearsals and standing executive advisory, so the estate holds its position instead of drifting back toward end of support over the next three budget cycles.
- Inputs
- Live estate and runbooks
- Monitoring and alerting feeds
- Vendor lifecycle updates
- Activities
- Monitoring and optimisation
- Quarterly restore rehearsal
- Lifecycle tracking
- Executive advisory
- Outputs
- Standing retainer
- Quarterly estate position
- Refresh forecast
Typical durationOngoing
Engagement Models
Three ways to hold the relationship
Managed service
Ongoing operation of the estate we designed, or of one we inherited and assessed first. Rehearsals are scheduled, not requested.
- Monthly, with a defined scope
- Quarterly restore rehearsal
- Named responder
Advisory retainer
Standing access to the technical position without a delivery commitment. For executives who need a straight answer before a board meeting.
- Fixed monthly hours
- Vendor proposal review
- Board-ready summaries
Boundaries
What we will not do
Commit a date before the rollback is tested
A cutover without a rehearsed way back is a gamble with someone else's operation. We will move the date rather than take that risk on your behalf.
Recommend a replacement we would profit from
We hold no resale margin. If the incumbent platform is supported, recoverable and efficient, the recommendation is to keep it.
Take work we cannot staff with senior people
If an engagement would need a delivery unit we do not have, we say so at qualification rather than discovering it during stage three.
Report a recovery time we have not measured
Recovery objectives in our documents are ones that have been executed under load. Anything else is an estimate and gets labelled as one.