Who this is for
If your cloud move has stalled, or spend and access have grown faster than governance, start with the foundations.
You probably need this if…
- A data-centre exit or hardware refresh is coming up.
- Cloud bills keep rising with no clarity on what's driving them.
- There's no tested disaster recovery plan for your cloud workloads.
- An audit or customer is asking exactly where your data lives.
- Multiple cloud accounts exist with no consistent governance between them.
- You're planning a new application and unsure which cloud model actually fits.
Where it applies
Probably not the right fit when…
- Stable workloads on fully depreciated hardware with no growth or compliance pressure.
- A single small application that a basic hosting plan already handles well.
Technical overview
Cloud outcomes are decided before the first workload moves: landing zone, identity, network and guardrails. We assess, build governed foundations, then migrate and modernise in dependency-based waves.
Technologies we work with
Reference architecture
Design principles
- Establish identity, network, policy, logging and recovery foundations before migrating production workloads.
- Choose rehost, replatform or refactor per workload using dependency, risk, business value and operating-cost assessments.
- Define availability, recovery point objective and recovery time objective for each workload and validate them through recovery exercises.
- Assess the DPDP Act 2023, DPDP Rules 2025, applicable commencement provisions, transfer restrictions and sectoral requirements for each workload.
Specification baseline
| Aspect | Specification | Why it matters |
|---|---|---|
| Identity | Use federated identity, multifactor authentication, least privilege and separate privileged administration | Identity controls govern access across cloud services and management planes. |
| Landing zones | Separate production, non-production, security and shared services using enforceable policy boundaries | Consistent foundations reduce configuration drift and limit blast radius. |
| Connectivity | Use private connectivity where justified, controlled ingress and egress, and documented DNS and routing | Predictable network boundaries improve security and troubleshooting. |
| Recovery | Assign workload-specific RPO and RTO tiers and test backups, dependencies and restoration order | Recovery targets must be demonstrated rather than assumed from availability claims. |
| Observability | Centralise security logs, workload telemetry, audit events and cost allocation | Operational and financial visibility support incident response and optimisation. |
| India data governance | Evaluate applicable privacy, transfer, sectoral and contractual obligations before committing to a data location | Data-residency requirements depend on the applicable legal and contractual context. |
Key design decisions
Migration strategy
Migration speed and limited initial application change are priorities and the target platform meets workload needs.
Improvements in scalability, resilience, delivery speed or managed-service adoption justify application changes.
Cloud footprint
Operational simplicity and standardisation outweigh the benefits of provider diversification.
Specific regulatory, resilience, acquisition or workload requirements justify additional integration and operating complexity.
Disaster recovery
Moderate recovery objectives justify maintaining a scaled recovery environment.
Short recovery objectives and business impact justify continuously available recovery capacity and data replication.
Sizing & calculations
RPO is the maximum tolerable data-loss interval; RTO is the target time to restore the required service after disruption.
An RPO of 15 minutes and RTO of 4 hours require restoration to a recovery point within 15 minutes of the incident and service restoration within 4 hours.
Monthly cloud cost = compute + storage + network transfer + managed services + support + applicable taxes and other charges.
A hypothetical monthly estimate of ₹80,000 compute, ₹25,000 storage, ₹15,000 networking and ₹10,000 services totals ₹1,30,000 before taxes and omitted charges.
Capacity planning must model average and peak demand, scaling limits, failover demand and workload growth.
A workload requiring 20 compute units normally and 35 at peak should be tested at peak load and expected failover load.
Standards, protocols & references
Handover documentation
- Cloud architecture and landing-zone diagrams
- Identity, network and policy configuration register
- Workload migration and dependency inventory
- Backup, RPO/RTO and recovery test results
- Security monitoring and alerting runbooks
- Cost baseline, budgets and operating procedures
What you get
- Predictable cloud spend through tagging, reservations and right-sizing
- Migrations sequenced by dependency, not guesswork
- A landing zone that scales without re-architecture
What we deliver
Strategy & advisory
- Readiness assessments and roadmap
- Landing zone and governance
- Risk, compliance, security review
Migration & modernisation
- Data-centre-to-cloud migration
- VM, database and storage migration
- Cloud-native architecture
Platform & infrastructure
- Kubernetes and container platforms
- Infrastructure as Code
- HA, DR and backup engineering
