What UN-Funded Projects Taught Us About Building for Scale

Over the past several years we've delivered platforms funded by UNDP, WHO, and the International Trade Centre — the Human Rights Commission Case Management System, the National GBV Reporting Platform, an electronic medical records pilot at National Hospital Kandy, and the National Trade Facilitation Monitoring Tool, among others. Each came with its own domain, but they shared a set of constraints that shaped how we build everything now, including commercial and government work that has nothing to do with a donor.

Sustainability planning starts before the grant does

Donor funding has a start date and an end date. The system doesn't. A platform that only works while the grant is active isn't a success by any reasonable definition — it's a well-documented failure with a delay built in. Every donor-funded system we deliver is designed with a maintenance and ownership plan for the period after the funding ends, because that's the actual test of whether the investment was worth making. It's the same discipline behind our standard multi-year maintenance contracts on nationally-funded work.

Multi-stakeholder governance is a design constraint, not a meeting problem

The National GBV Reporting Platform connects five separate state institutions — police, CCID, NCPA, MoWCA, and SLCERT — into one case-management backbone. That's not just a stakeholder management challenge during the build; it shapes the actual architecture. Access control has to reflect five different institutional mandates. Case routing has to match real inter-agency escalation paths, not a simplified version of them. When a system needs to serve multiple institutions with different authority, different data sensitivity requirements, and different reporting obligations, you can't retrofit that governance model after the schema is already fixed — it has to be a first-class design input.

Documentation is part of the deliverable, not an appendix

Donor-funded projects are audited, evaluated, and frequently handed to a different implementation partner down the line. That means documentation — architecture decisions, data dictionaries, handover procedures, training materials — has to be complete enough for someone outside the original team to operate and extend the system. We treat this the same way on every national platform now, donor-funded or not, because institutional staff turnover makes it just as necessary on a ministry-funded system as on a UNDP-funded one.

Interoperability has to be proven, not just claimed

The National EMR pilot at National Hospital Kandy, funded by WHO, required a Master Patient Index that could actually enable data sharing between the hospital and primary care institutions — not a theoretical integration point, but real interoperability, validated against real clinical workflows and real user needs assessments before rollout. Similarly, the National Trade Facilitation Monitoring Tool has to produce data that's usable by policymakers and donors evaluating Sri Lanka's WTO Trade Facilitation Agreement commitments — which means the data model has to match how international stakeholders actually consume that information, not just how it's convenient to store internally.

Security and privacy get treated as launch blockers, not launch risks

Systems built for sensitive domains — human rights case handling, gender-based violence reporting, anti-corruption reporting for Transparency International Sri Lanka — don't get the option of shipping with security as a fast-follow. Encrypted communication channels, audit trails, and access controls have to be validated before a single real case enters the system, because the population these systems serve is often the population most at risk if something goes wrong.

Why this matters beyond donor-funded work

None of these constraints are unique to international development funding — they're just enforced more rigorously there, because the accountability structures (donor evaluations, multi-agency governance, public sector audits) leave less room to skip them. We've found that holding every project to that standard, government or private, donor-funded or commercially funded, is the difference between a system that survives its first staff turnover and one that doesn't.