Platform vs. build-from-scratch: the honest cost of in-house infrastructure

Every operator we meet has modelled the build-it-in-house scenario. Engineering teams quote 12 to 18 months and a headcount. The actual lived experience is 24 to 36 months and a second headcount — plus the features that never got built because the team was busy keeping the platform up.

Where the time actually goes

• Identity and auth: 3–6 months to get OIDC, SSO, RBAC and audit trails right — before any business logic.
• Multi-tenancy: 2–4 months of schema design, migration tooling, and tenant-scoped queries.
• Payments and financial reporting: 3–9 months including vendor integration, reconciliation, and regional tax models.
• Compliance: 6–12 months for SOC 2 / ISO 27001 readiness if done alongside the build.
• Ops tooling: 2–4 months for observability, release process, and support runbooks.

None of that produces differentiating product value. It produces the table stakes that let the differentiating product exist.

What a modular platform compresses

FlyttGo ships all of the above as independent modules — Identra for auth, Payvera for payments, Ledgera for financial ops, Civitas / EduPro / Transify / Workverge for domain logic. Operators pay for the modules they use and focus their teams on the 20% that is actually unique to their programme.

When to still build it yourself

If your differentiation genuinely lives at the infrastructure layer — novel clearing rails, new identity protocols, or core algorithms that need to ship from first principles — build. For the 95% of programmes where the differentiation lives in product, policy, or go-to-market, the modular platform wins.