What we move
- On-prem → cloud. Workload-by-workload assessment, dependency map, target-state design, IaC scaffold, networking, IAM blueprint, cutover sequencing. Common shapes: VMware / KVM workloads to EC2 / GCE / Azure VMs, on-prem databases to RDS / Aurora / Cloud SQL / Azure SQL, on-prem data warehouses to Redshift / BigQuery / Snowflake / Databricks.
- Cloud → cloud. AWS ↔ GCP ↔ Azure ↔ OCI, whole-footprint or workload-level. The rare but real cases when an existing cloud just does not fit the application or finance shape any more.
- Cloud-native rebuild. Lift-and-shift is sometimes the wrong answer. When workloads need re-architecting to actually pay for the move (cold storage tiers, serverless replacements, managed-service adoption), we re-architect alongside the migration. Done with a written architecture in the Appendix A so it is not a moving target.
- Landing zone and account / subscription topology. AWS Control Tower / Landing Zone Accelerator, GCP Foundation, Azure Landing Zones — set up to your security and FinOps posture, not the vendor reference architecture verbatim.
- Disaster recovery and business continuity. RPO / RTO budgeting per workload, cross-region replication topology, backup automation, restore-test scheduling, runbook authoring.
What you receive
- Written cloud migration plan — workload inventory, dependency map, complexity tier per workload, target-state architecture, IaC scaffold approach, wave plan with sequencing constraints, 3-year TCO model.
- Reconciliation evidence per wave — for any data move, row counts, schema parity, query-result diffs on a target sample. Same deterministic harness as our data-platform migrations.
- Landing zone IaC scaffold — Terraform organized by environment, modular, baseline security guardrails wired up, multi-account governance posture documented.
- Signed Appendix A — scope contract with the 90-day Scope Warranty.
How we price cloud engagements
Every cloud engagement starts with the same 10-day Discovery Sprint (₹2 – 3 L / $3 – 5k, 100% upfront). At day 10 you receive a written diagnostic, a recommended target-state, a wave plan with effort range, and a fixed Execution quote. Execution is milestone-billed (30% upfront / 40% mid / 30% on hypercare exit). Sign Execution within 60 days of Sprint delivery → 5% Combo discount applies.
What is out of scope
We do not handle network procurement (telco / interconnect contracts), end-user training programs, hardware decommissioning logistics, or vendor co-fund / migration grant paperwork beyond providing the technical artifacts your procurement team needs. If those are required we hand off cleanly — we do not bundle billing for work that is not cloud engineering.
Why this is not just lift-and-shift
The pattern we see across SI-driven cloud migrations: a 100% lift-and-shift that does not pay for itself, no FinOps work post-cutover, and an angry CFO at month six. We refuse to ship that shape. The Discovery Sprint deliberately surfaces which workloads should be retired (typically 15-25% of inventory), which need re-architecting before move (5-15%), and which are genuinely lift-and-shift candidates. Most of our cloud engagements end with a smaller cloud footprint and a lower bill than the source.
When multi-cloud is the wrong answer (and when it is right)
Multi-cloud is what executives ask for after one cloud-vendor lock-in scare. It is usually the wrong answer — the operational tax is real, the cross-cloud networking math is brutal, and the security surface area doubles. We tell you that up front. The cases where multi-cloud is genuinely right: a regulator requires it, a single-cloud incident is an existential risk for your business, or your workload mix has genuinely different cloud-cost profiles per workload. None of these apply to most of our clients; if they apply to yours, we design for it cleanly.
FAQ
Which clouds do you actually work with?
AWS, GCP, Azure, and OCI as production targets. We have built migration packs for AWS ↔ GCP, AWS ↔ Azure, and on-prem to AWS / GCP / Azure. Multi-cloud designs are a thing we do but tell you when not to.
How long does a cloud migration take?
Discovery Sprint is 10 business days fixed. Execution ranges from 8 to 24 weeks depending on workload count, complexity, and how much of the estate is genuinely worth migrating versus retiring or modernizing in place.
Do you handle the database migration as part of cloud migration?
Yes — most cloud migrations include a data layer move (Oracle / SQL Server / DB2 → Postgres or Aurora; Redshift / BigQuery / Snowflake moves). The data layer is explicit in Appendix A so it is never the surprise that breaks timeline.
What about FinOps after we are on the new cloud?
A FinOps audit is a separate 10-day engagement we ship after cutover. Most clients book it 30 days post-cutover when the bill has stabilized — that is when the savings are most surfaceable. See FinOps and Cost Reduction.