Skip to content
Infrastructure

Novaworks re-architected for autoscaling and cut infrastructure cost by 33%

A 500-person platform engineering company ran its core products on always-on servers sized for peak traffic. Moving to a containerized, autoscaling architecture with a leaner DR posture cut infrastructure spend from ₹38L to ₹25.5L a month.

  • Enterprise software and platform engineering
  • 500 people · Large
  • Noida
  • 10 weeks

Monthly spend

₹38L₹25.5L
Saved / month
₹12.5L
Reduction
33%
Payback
5 weeks

−33%

Infrastructure cost

₹1.5Cr

Annual saving

−60%

Overnight capacity

Unchanged

RTO commitment

The challenge

Novaworks' platform had been designed a decade ago for a fixed fleet of servers. Traffic varied fivefold between night and day, but capacity stayed constant.

A hot-standby disaster recovery site mirrored production in full, doubling the cost of resilience.

What we reviewed

  • Architecture of 6 core products and 140 services
  • Traffic patterns and capacity utilization over 6 months
  • Disaster recovery requirements and RTO/RPO commitments
  • Licensed operating systems and middleware
  • Caching, queuing and storage tiers

The approach

  1. 1

    Containerize the stateless tier

    Stateless services moved to a managed container platform with horizontal autoscaling tied to real demand.

  2. 2

    Scale with the day

    Capacity now follows traffic, dropping by 60% overnight and on weekends.

  3. 3

    Right-size disaster recovery

    Hot standby was replaced with a pilot-light design that still meets the contracted recovery objectives.

  4. 4

    Reduce licensed middleware

    Open-source equivalents replaced two licensed middleware components, and OS licences dropped with the smaller fleet.

Before and after

Monthly cost by line item, in Indian rupees.

Novaworks — monthly infrastructure spend

Line itemBeforeAfterSaving
Always-on application servers₹16,50,000₹9,40,000−₹7,10,000
Managed container platform (autoscaling)—₹2,10,000+₹2,10,000
Load balancing and networking₹4,20,000₹2,96,000−₹1,24,000
Caching and queue tier₹3,10,000₹2,10,000−₹1,00,000
Storage volumes₹5,40,000₹3,80,000−₹1,60,000
Disaster recovery standby₹6,20,000₹3,40,000−₹2,80,000
Licensed OS and middleware₹2,60,000₹1,70,000−₹90,000
Total₹38,00,000₹25,46,000₹12,54,000 (33%)

Monthly spend over time

₹1.5Cr saved per year

Two months before the engagement, then six months after the plan was delivered.

Timeline

  1. Weeks 1–3

    Architecture review

    Services, traffic and DR requirements mapped.

  2. Weeks 4–6

    Target design

    Container platform and DR designs reviewed with architects.

  3. Weeks 7–8

    Pilot

    Two products migrated and load-tested.

  4. Weeks 9–10

    Roadmap

    Phased migration plan for the remaining products delivered.

“Brian and the team understood a ten-year-old architecture faster than some of our own engineers. The design they proposed was pragmatic, and the savings are larger than our original target.”
GMGaurav MalhotraChief Architect, Novaworks

Find out what your agency could save

Book a 20-minute call. We'll show you where the quickest wins usually are for a team like yours, with no data required and no obligation.

  • Free 20-minute assessment
  • NDA before any data is shared
  • Fixed-scope quote, 30/70 payment terms
  • 7 days of post-delivery support
Chat with us