Scale-Out Cloud Migration of 2,000+ Servers to GCP
Led the team migration of Ericsson's Citrix virtual infrastructure — 2,000+ production servers — from on-prem to Google Cloud Platform, one of the company's largest public cloud modernization initiatives, with hardened hybrid security and automated database operations.
Problem
Ericsson's Citrix virtual infrastructure ran on aging on-prem hardware with no clear path to scale. 2,000+ production servers were expensive to maintain, hard to resize, and a single point of failure for tool uptime. The company needed one of its largest public cloud modernization initiatives executed without disrupting the internal tools that global teams depended on daily.
Approach
Ran discovery and assessment across the 2,000+ server estate to map dependencies before touching production, so migration batches could be planned around application affinity.
Built the GCP landing zone first — networking, identity, and quotas — then executed a scale-out Citrix migration in waves rather than a big-bang cutover.
Collaborated with security teams to harden the resulting hybrid cloud environment during migration, treating security as a migration track rather than a post-migration review.
Automated database management workflows alongside the migration so post-cutover operations did not inherit the manual maintenance burden of the old estate.
Outcome
-30%
Infrastructure cost
estimated reduction after migrating 2,000+ production servers from on-prem to GCP
99.9%
Uptime SLA
three nines on migrated tools, up from the prior on-prem reliability baseline
-40%
Maintenance time
database management workflows automated, freeing roughly 8 hours per week for the team
Pipeline
Evaluation
The 30% cost reduction is an estimate built from the difference between on-prem hosting costs (hardware refresh, power, licensing) and the GCP consumption plan post-migration; the 99.9% SLA is measured from tool availability after cutover. Both were tracked per wave during the migration, which is how the team could prove the program was working before the final waves landed.
What I'd Do Differently
I would have started the security hardening track even earlier. Collaborating with security teams mid-migration worked, but weaving their requirements into the landing zone from week one would have removed rework in later waves. I would also standardize the per-wave cost and uptime reporting into a single dashboard from the start — the numbers existed, but they lived in too many spreadsheets before I consolidated them.
Tech Stack
Related Projects
Enterprise-grade automated threat intelligence platform combining a FastText NLP classifier, structured ML models, and algorithmic distance rules to identify deceptive URLs and brand spoofing for client security teams at NetSTAR Global.
Distributed Medallion Lakehouse integrating cross-ministry CSR, groundwater, and education datasets using Apache Spark and Iceberg for panel regression analysis of Indian state-level policy outcomes.
End-to-end ELT streaming pipeline that ingests real-time Twitch viewership data through Apache Kafka, warehouses in Snowflake, and transforms with dbt for Looker Studio dashboards.