"SkyliftAI MigDB was instrumental for our on-premise database server migration to Azure Managed RDS. The platform delivered confidence, control, and a level of operational safety we had not seen in prior migration efforts."
University College London (UCL), one of the world's leading research universities, set out to modernise the database backbone supporting its academic, research, and administrative operations. With a portfolio of 200 production SQL Server instances running on-premise, UCL's Infrastructure Support Team needed to migrate the entire estate to Microsoft Azure Managed RDS in the London region — without interrupting business applications relied on by tens of thousands of students, faculty, and staff.
Partnering with SkyliftAI, UCL deployed the MigDB platform to orchestrate the migration end to end. All 200 databases were successfully transitioned to Azure in just 48 hours with zero data loss, zero downtime, and no perceptible impact to users. The programme finished 60% faster than comparable migrations, reduced migration costs by 40%, and lowered total cost of ownership by 25%.
University College London is a multidisciplinary research university based in London, United Kingdom. Its Infrastructure Support Team is responsible for the resilience, performance, and security of the data platforms underpinning teaching, research computing, finance, HR, and student-facing services. The team operates under strict service-level commitments where unplanned downtime is not tolerated.
UCL's legacy on-premise SQL Server estate had reached a point where hardware refresh cycles, licensing economics, and scalability constraints were limiting innovation. Migrating to Azure Managed RDS in the London region was the strategic choice — but the operational requirements were uncompromising.
Migration windows had to be invisible to end users. Any disruption to academic or administrative workloads was unacceptable — tens of thousands of students, faculty, and staff depend on these systems continuously.
200 SQL Server databases needed to be moved into the Azure London region, many with interdependencies and large transactional workloads — requiring a platform capable of orchestrating parallel migration waves without manual stitching.
The business required bi-directional replication post go-live so that, in the event of any anomaly, the team could fail back to source instantly with full data integrity — removing the fear of irreversibility from a 200-database cutover.
Multiple databases had to be migrated in parallel using one platform, removing the operational risk and complexity of stitching together point tools — with a single pane of glass for monitoring, alerting, and troubleshooting across all waves.
SkyliftAI MigDB was deployed as the unified migration platform across UCL's entire SQL Server estate. The platform combined enterprise-grade change data capture, intelligent parallelisation, and bi-directional replication to meet every operational guardrail the Infrastructure Support Team set.
Continuous change data capture kept source and target in lockstep, so business applications never had to pause writes during cutover. Every transaction that occurred during the migration was captured and applied to the target with full fidelity.
Optimised concurrent streams moved baseline data into Azure at maximum throughput, dramatically compressing the migration timeline — enabling all 200 databases to complete within the 48-hour window that the programme required.
Post go-live, changes replicated both ways between Azure and on-premise, giving stakeholders a tested, push-button rollback option and removing the fear of irreversibility from the largest database migration in UCL's history.
All 200 databases were orchestrated through one console, with consistent monitoring, alerting, and troubleshooting workflows across the entire wave — eliminating the complexity of managing multiple point tools simultaneously.
The outcome exceeded the original programme commitments on every dimension that mattered to UCL's leadership: speed, safety, and economics.
200 production databases migrated to Azure Managed RDS in 48 hours.
Zero data loss and zero downtime across the entire estate — no measurable impact to business applications or end users.
60% faster migration compared with the team's prior benchmarks for projects of similar scale.
40% reduction in overall migration cost versus the originally budgeted alternative.
Programme delivered exactly as planned, with onboarding to go-live completed within four weeks.
75% faster onboarding cycle and 25% reduction in total cost of ownership going forward.
The success of the migration was underpinned by the breadth and maturity of the SkyliftAI MigDB platform. The following capabilities were activated for UCL.
Continuous CDC and synchronised cutover kept business applications fully available throughout the transition — tens of thousands of students, faculty, and staff experienced no disruption.
Encryption in transit and at rest, role-based access, and audit trails aligned with UCL's information governance standards and higher-education data compliance requirements.
An intuitive console allowed the Infrastructure Support Team to plan, execute, and validate waves without deep scripting — accelerating the team's time to competency significantly.
Real-time dashboards exposed lag, throughput, and integrity metrics across all 200 databases concurrently — giving the team complete visibility from a single pane of glass throughout the 48-hour window.
Granular logs and guided diagnostics shortened mean time to resolution during validation windows — a critical capability when managing 200 databases across a compressed migration timeline.
A consistent control plane across Azure and other targets future-proofed UCL's data platform strategy — ensuring the migration investment delivers value beyond the current Azure programme.
Parallel workers and elastic compute absorbed peak transfer loads without manual tuning — enabling the 200-database wave to complete in 48 hours without performance degradation or backlog accumulation.
Templated runbooks turned a complex 200-database programme into a repeatable, auditable workflow — reducing human error risk and providing a reference pattern for future migration waves.
For a research-led institution like UCL, the database tier is not a back-office concern — it is the foundation of teaching continuity, research productivity, and operational trust. SkyliftAI MigDB allowed the Infrastructure Support Team to retire a complex on-premise estate, modernise on Azure Managed RDS, and bank meaningful operational and financial gains, all while keeping the promise of "no downtime" to the business.
The migration is now a reference programme within UCL for how large-scale data platform transformations can be delivered quickly, safely, and economically.
Ready to modernise your database estate with zero downtime and zero data loss? Talk to a SkyliftAI MigDB expert today.