Solution · Modernization

A multi-year program
the board can actually read.

Two hundred and forty-seven applications. Three datacenters. Twelve quarters. Panaptico runs the portfolio as a live graph — every app classified, every wave overlapped by design, every domain strangled instead of rewritten, every outcome traced back to the budget that paid for it.

Portfolio Canvas· MOD-VMW-AWS-0003 · Quarter 7 of 12
247 apps · 4,180 VMs · 3 datacenters

From · on-prem

VMware · 3 datacenters

VMs
4,180
Applications
247
Annual infra
$18.2M
Utilization
28%

Program state

112 / 247 apps

Rehost · 54Replatform · 71Refactor · 42Rebuild · 23Retire · 31Retain · 26

Migrated

112

In flight

95

Queued

40

To · cloud

AWS · 2 primary + 1 DR

Containerized
EKS
Serverless
Lambda
Projected run-rate
$9.1M
Target utilization
64%

Current quarter

Q7 of 12 · 58% done

DC Dallas

offline · Q6 ✓

Cost run-rate

$18.2M → $11.4M

Board sign

Q4 2026 · next

The gap

The three-year program that always ends in its fifth year.

01

Portfolio is a word that hides details

‘We’re moving 247 apps to the cloud.’ Eighty of them are duplicates nobody’s admitted to. Twenty belong on-prem for regulatory reasons. The spreadsheet doesn’t know.

02

You can’t freeze the business for three years

The product keeps shipping while the foundation gets rebuilt. Every wave ships into live traffic. Modernization without a strangler-fig plan becomes a slow rewrite that the market eats alive.

03

The board signs the budget, not the design

Year one is landing zones. Year two is moving the easy apps. Year three is the hard architectural calls — and by then the sponsor is gone, the team has rotated, and nobody remembers why.

Portfolio disposition

Every app gets a verdict. Signed, not vibed.

Panaptico classifies the full portfolio against the 6-R framework — rehost, replatform, refactor, rebuild, retire, retain. Each decision is bound to the systems still depending on it.

Rehost54

Lift-and-shift · minimal change

  • internal wiki
  • legacy CRM
  • 3 batch ETLs
  • licensing portal
Replatform71

Lift-tinker-shift · managed services

  • billing gateway → RDS
  • invoice service → EKS
  • 2 search indexes → OpenSearch
Refactor42

Re-architect within the same bounded context

  • catalog API · stateless split
  • notifications · SQS fan-out
  • reporting · query-side rewrite
Rebuild23

Decompose · cloud-native from scratch

  • checkout monolith → 6 services
  • auth broker → Cognito + custom
  • order flow · event-sourced
Retire31

Decommission · duplicate / end-of-life

  • 3 duplicate analytics stacks
  • EOL financial reporter
  • Lotus Notes archive
  • shadow SharePoint
Retain26

Stay on-prem · latency / regulatory

  • trading engine · sub-ms latency
  • regulated records vault
  • factory-floor SCADA bridge

Program roadmap

Twelve quarters. Five streams. One graph.

Foundation first, then waves overlapping by design. Every stream is tied to signed outcomes and to the datacenter retirements they unlock. The current quarter marker advances with the graph, not with the slide deck.

Q1Q2Q3Q4Q5Q6Q7Q8Q9Q10Q11Q12

Foundation

Landing zones · IAM · CI/CD · observabilitydone

Wave 1 · rehost

54 apps · lift-and-shiftdone

Wave 2 · replatform

71 apps · managed DB + K8s

Wave 3 · refactor

42 apps · bounded-context split

Wave 4 · rebuild

23 apps · event-sourced · serverless

Retire · decom

Dallas DC · offlinedone
North VA DC · offline
Final cutover + records sealed
FoundationRehostReplatformRefactorRebuildRetirecurrent · Q7

Strangler-fig decomposition

One monolith. Six domains. Zero rewrite-in-a-vacuum.

The checkout monolith couldn’t be rebuilt as a single cutover. Panaptico decomposed it into bounded contexts, routed traffic domain by domain, and preserved the transaction graph across the seam.

System under modernization

checkout-monolith · 2014 · Java 8 · MSSQL

50% decomposed · 3 of 6 domains live
D1done

Catalog

Rehosted · Q3 · RDS Aurora · read replicas

D2done

Pricing

Refactored · Q5 · stateless · API Gateway

D3done

Inventory

Refactored · Q6 · event-sourced warehouse

D4active

Payment

Rebuilding · Q7 · Lambda + DynamoDB · PCI scope re-scoped

D5queued

Orders

Queued · Q9 · Step Functions · saga pattern

D6queued

Fulfillment

Queued · Q10 · EKS · warehouse plugin SDK

Traffic · monolith vs new services

Q3
100% · 0%
Q4
82% · 18%
Q5
58% · 42%
Q6
34% · 66%
Q7
16% · 84%
Q8
0% · 100%
Monolith New services

Outcome ledger

What the board is actually buying.

Every quarter, the graph reports against the outcomes the budget was signed for — not velocity, not activity. Lead time, cost, resilience, capability. If a wave doesn’t move the number, it doesn’t ship.

Feature lead time

↓ better

Was

14 days

Now

2.3 days

Target

1 day

deploy frequency 2/mo → 23/wk

Annual infra cost

↓ better

Was

$18.2M

Now

$11.4M

Target

$9.1M

DC Dallas offline · 64% utilization target

MTTR · service incident

↓ better

Was

4h 12min

Now

47 min

Target

15 min

observability stack · automated runbooks

DR recovery (RTO)

↓ better

Was

8 hours

Now

22 min

Target

<15 min

multi-region active · chaos-tested Q6

Datacenter footprint

↓ better

Was

3 DCs

Now

2 DCs

Target

1 DC (Amsterdam · retain)

regulatory hold on EU workloads

Cloud-capable engineers

↑ better

Was

31%

Now

58%

Target

85%

curriculum + pairing · Platform team sponsors

A three-year program
that remembers why it started.

Panaptico binds every wave, every decision, every decommission to the outcomes the board signed for — so the program is legible from quarter one to quarter twelve.

← M&A integration·Related: Post-go-live continuity