By Janel Chadiarova, Full-Stack Data Professional
What to do when the data you need doesn’t exist — and the system it should live in was never built.
In emerging economies, digital transformation often begins not from a lack of data, but from its fragmentation and uneven maturity across regions and sectors, creating a paradox where ambition outpaces the systems needed to support it. Ambitious strategies for innovation and sustainability are launched every year — yet the essential data to drive them doesn’t exist. There are no interconnected databases, no reliable baselines, and sometimes not even a shared definition of what should be measured.
In these contexts, data-driven decision-making is not just a technical challenge but a question of leadership and design. The task is not to automate what already works elsewhere, but to build systems from zero: connecting fragmented information, establishing trust in the numbers, and turning scattered evidence into intelligence that investors, policymakers, and citizens can actually use.
This article outlines a practical framework for designing data systems in low-infrastructure environments — combining creativity, governance, and cross-disciplinary thinking. It is drawn from real-world implementation, including the creation of Kazakhstan’s first renewable-energy investment dashboard.
1. Creativity First: Design the Data You Need
In system-poor settings, the starting point is often conceptual. The first question is not what data we have, but what we need to know to solve the problem or enable the investment.
In the absence of formal databases, the role of a data architect becomes part strategist, part product designer. The objective is not to replicate existing data models but to identify proxy indicators, local intelligence, and workarounds that can be structured into meaningful datasets.
This step is both creative and analytical: it asks not just what data is missing, but what can be constructed to fill the gap. Designing data from scratch requires imagination — envisioning the information that could exist — and discipline to ensure it is measurable, traceable, and valuable.
2. Find the Right Sources: Fragmented, Informal, Overlooked
Data almost always exists — just not where or how we expect it.
It may reside in regulatory filings, project tenders, scanned PDFs, ministerial reports, or even in the knowledge of local experts.
In emerging markets, the most valuable insights often come from non-traditional sources: public records, development finance reports, local news archives, or NGO field surveys. Extracting and validating this information requires a cross-functional team fluent in both policy and digital forensics.
The challenge lies in recognising the latent value of overlooked data and converting it into usable intelligence. Each fragment, once structured and verified, becomes part of a mosaic that can inform larger systems.
3. Use Diverse Collection Techniques: Manual, Automated, and Human-Verified
Once potential data sources are identified, the next step is operationalising the collection.
In the absence of APIs or clean datasets, a hybrid strategy becomes essential — one that balances speed and accuracy:
- Manual extraction from unstructured sources such as PDF reports, spreadsheets, or scanned tables.
- Automated scraping from publicly available websites — including regulatory portals, tenders, or data repositories.
- Batch updates from structured platforms, where available, to ensure continuity.
- Human-in-the-loop quality control, ensuring contextual accuracy and integrity.
This blended model allows teams to move fast where automation works and apply human judgment where nuance matters. The process itself becomes a bridge between traditional research and digital transformation.
4. Build Fit-for-Purpose Infrastructure: Scalable, Not Fragile
In low-infrastructure environments, building a resilient data pipeline means prioritising function over form.
The system does not have to be complex — but it must be maintainable, context-appropriate, and capable of growing over time.
Cloud-native architectures may be ideal in theory, but in practice, a well-structured shared database or even a collaborative spreadsheet with version control and a simple dashboard interface can deliver faster and more lasting impact.
What matters is that the system supports:
- Regular and transparent updates
- Geographic or thematic filtering
- Explicit access permissions and governance
- A structured path toward scale or integration with national systems
Infrastructure, in this context, is as much about governance as it is about the tech stack. Sustainable systems are not only technically sound but also institutionally owned and trusted.
5. Communicate Value Through Storytelling: Data Alone Isn’t Enough
Even the most accurate data has a limited impact if it is not understood. Once a system is built, the final — and often most strategic — step is narrative translation: turning rows and columns into actionable intelligence.
This may take the form of:
- Interactive dashboards tailored to different stakeholder groups
- Insight briefs summarising key trends and their implications
- Visual summaries for public engagement or investor communication
- Performance trackers aligned with national development or targets
Storytelling transforms data from a static resource into a dynamic decision-making tool. The aim is not just to inform, but to inspire action — to make the data speak in a language that leaders, investors, and citizens can all understand.
Case Study: Kazakhstan’s First Renewable-Energy Investment Dashboard
A concrete application of this framework was the development of Kazakhstan’s first publicly available, centralised renewable-energy investment dashboard, designed specifically for use by investors and policy stakeholders.
Before this project, no centralised database existed for the country’s green-energy infrastructure. Investors, policymakers, and multilateral stakeholders lacked a single, reliable view of the sector.
In response, we designed and launched a tool that aggregated and visualised key information, including:
- Type and technology of each renewable plant (solar, wind, hydro, biomass)
- Installed capacity (MW)
- Commissioning year
- Investor or sponsor entity
- Geographic location
- Regulatory or licensing status
The data were assembled from a combination of government sources, development bank publications, public tenders, and manual verification. Much of it had never been publicly structured before.
The platform was designed to be bilingual (English and Russian), mobile-friendly, and periodically updated. It quickly gained traction among ministries and international investors and was referenced in policy and project discussions.
It demonstrated that even in the absence of formal systems, data can serve as a common language for cooperation among government, business, and civil society.
Conclusion: From Information Gaps to Strategic Intelligence
Designing data systems in environments without digital infrastructure demands more than technical know-how. It requires:
- Creativity to define what data is truly needed
- Rigour to source and validate it
- Tactical tools to collect and clean it
- Lean infrastructure to manage and present it
- Strategic communication to ensure it informs real-world action
The ultimate value of data lies not in its existence but in its ability to drive brighter, fairer, and faster decisions.
In places where no systems exist, those who can design them hold not just a technical advantage — but a strategic one.
