Identity Data in Digital Public Infrastructure
Infrastructure treats identity as fixed. People aren't
Every identity system, whether it's patient records, student records, civil registration, or national ID, eventually has to handle a name that changes, a gender marker that's recategorised, a record duplicated across sites, a person who doesn't fit the schema anymore. I've seen this problem from the ground up, building offline data systems for 20,000 students in Thailand-Myanmar refugee camps, and stabilising a provincial OpenEMIS rollout in South Africa. My PhD then let me study why these failures happen at scale. Right now, I'm the identity data specialist on the early design of a DHIS2-based identity tracking system for a large international NGO, exactly the kind of high-stakes, high-complexity identity work I specialise in. I bring both field-tested engineering and research-level understanding of identity infrastructure to deduplication, record linkage, and identity system architecture.
PhD Digital Public Infrastructures, University of Oslo
DHIS2 Contributor (Python, React)
Research Affiliate, University of Edinburgh
Best Paper Award, Scandinavian Conference on Information Systems, 2025
Services
Identity Linkage, Deduplication, and Interoperability Design
Design and review of record linkage, deduplication strategy, and identity data flows across patient ID systems, civil registration, and vital statistics (CRVS), and student/beneficiary records. Grounded in hands-on OpenEMIS deployment experience, current work as identity data specialist on a DHIS2 identity-tracking implementation, and close knowledge of DHIS2's architecture as a codebase contributor. Following MOSIP and OpenCRVS closely, since these platforms are directly relevant to my specialism.
Identity System Architecture Review and Selection
Independent, vendor-neutral advice on whether DHIS2, OpenEMIS, MOSIP, or a hybrid approach fits an organisation's identity data needs, technical capacity, and governance structure. Grounded in hands-on OpenEMIS deployment experience and deep familiarity with DHIS2's codebase and design decisions.
Identity Governance and Readiness Advisory
Support thinking through how an organisation defines, governs, and changes identity categories such as gender, disability status, legal status, and nationality, before a system is built or scaled, informed by PhD research into how identity infrastructure succeeds or fails at the institutional level.
Hands-on Implementation
Production-level engineering in Python, Java Spring Boot, React, and PHP, applied to identity and record-management systems specifically. Deployment support experience includes managing an outsourced development team on a live OpenEMIS rollout to reduce bugs and stabilise the system, plus direct contributions to the DHIS2 codebase as a research-affiliated developer.
How Engagements Work
Most engagements start with a systems review call for 30-45 minutes at no cost, where we work out whether this is actually a fit before either of us commits to more. From there, engagements usually take one of two shapes: a focused advisory piece (architecture review, governance readiness, platform selection) delivered over two to six weeks, or ongoing implementation support alongside your team, scope to the engagement.
I work vendor-neutral and tell you early if something's outside what I can genuinely help with. I'd rather point you elsewhere than take on work I'm not the right fit for.
Who It's For
Implementing Partners · NGOs · CRVS and National ID Programmes · Organisations Building Towards DPG Standards
Implementing Partners and NGOs
Managing patient, student, or beneficiary records in the field, especially where deduplication or fragmented systems are the core problem.
CRVS and National ID Programmes
Working on civil registration or national identity infrastructure and want someone who can speak to both the governance side of identity categories and the underlying code.
Ministries and Global Funders
Partnering with senior implementers on larger engagements. Get in touch to discuss.
Working Globally
Edinburgh is home base. Click a location marker to see the work.
Click a marker above to see details about the work there.
Map data: Al MacDonald / Fritz Lekschas, CC BY-SA 3.0
Case Studies
The work below spans public sector, health, humanitarian, and education systems, the same interoperability, data quality, and system-linkage problems that come up across DPG Alliance-aligned infrastructure, applied in the field.
SMRU - Data Quality (OpenValid)
The Problem:
Shoklo Malaria Research Unit, a medical research centre on the Thailand-Myanmar border serving displaced migrants from Myanmar, ran its data management on MS Access databases with unreliable validation, leading to medical data entered outside normal ranges.
What I Did:
Built OpenValid, a customisable VBA validation module medical staff could configure themselves across all their databases. Gathered validation standards directly from doctors and medical staff, trained the IT team so the system could be maintained locally, and trained data entry staff on the new prompts.
What Changed:
A measurable reduction in data quality issues, giving the research teams cleaner data to work from (2018-2019, Developer and End-user Training)
World Education - Rural Education Tracking System - Data Collection
The Problem:
A regional education authority in Myanmar, on a project lead by World Education and funded by USAID, oversaw primary and secondary education for 180,000 students across 1,700 schools, most of which were only reachable on foot. Student data tracking was challenging because it took place with unreliable power and no mobile signal, tracked entirely on paper.
What I Did:
Designed and implemented a digitised data collection system running on 100 thin-client laptops built to work reliably despite unstable power. Used a train-the-trainer approach, training a core group within the education authority who then trained data entry and collection staff, specifically to work around language and cultural barriers and to make sure the system would outlast my involvement.
What Changed:
Digitised, donor-ready reporting for USAID and World Education, and a locally-owned training structure that didn't depend on an outside consultant to sustain it. (2018, Development and Implementation)
OpenEMIS - Digital Transformation
The Problem:
The Eastern Cape Ministry of Education in South Africa, on a project funded by UNESCO, needed to roll out OpenEMIS, an open-source education information system, entirely remotely during the COVID-19 pandemic, with a development team based in South Asia and an implementation team in South Africa.
What I Did:
Acted as a technical consultant: trained the outsourced software development team to build custom features with strong unit and system test coverage, supported the system administration team on server management, and wrote extensive documentation for every user role, all delivered over Zoom under travel restrictions.
What Changed:
A fully trained implementation and admin team, custom features built to a tested standard, and documentation that let the ministry sustain the system without ongoing outside support. (2021, Technical Consultant: Training and Documentation)
M-FUND - Data Pipeline
The Problem:
M-FUND, a social enterprise in Thailand and Myanmar funded by UNICEF providing universal health insurance coverage, needed to collect and analyse participant data in regions with no mobile internet signal, on low-cost hardware.
What I Did:
As Technical Lead, built a Bring Your Own Device (BYOD) offline data pipeline, with data collected offline and auto-uploaded once a device reached signal, processed via an Offline Analytical Processing (OLAP) approach on low-cost servers, feeding a customisable dashboard for donor and management reporting.
What Changed:
Enabled daily reporting connecting donor funding to real outcomes, and data-driven strategic and financial planning for the organisation's leadership, a system that ran reliably on hardware most vendors would consider too constrained. (2019-2021, Technical Lead in change of software delivery)
Save The Children - Refugee Camp Education Data System
The Problem:
Save the Children needed an education data management system for refugee camps in Thailand and Myanmar, where connectivity couldn't be assumed and the system had to work entirely offline.
What I Did:
Architected and built the system form the ground up, implementing it in direct collaboration with the local education authority so it fits the realities of camp-based service delivery rather than assuming standard infrastructure.
What Changed:
Gave the local education authority a working, fully offline data system covering around 20,000 students living in the refugee camps, with data collection that didn't depend on connectivity the camps didn't reliably have.
The Three Seams
Identity data rarely fails at the centre. It fails at the seams: between systems that were never built to share a common notion of "who this person is", between a well-designed identity system and the institution that has to govern and fund it, and between a record made once and a person whose name, gender marker, legal status, disability status, or life circumstances don't stay still. I've seen this pattern play out directly in the field, across patient records, student records, and civil registration, and I've studied it in depth through my PhD.
Technical
DHIS2, MOSIP, OpenEMIS, Civil Registration: Patient IDs, civil registration, and national identity platforms rarely speak to each other cleanly. Without deliberate linkage design, systems like these end up duplicating records instead of connecting them.
Institutional
Procurement, Government-fit: Even a well-designed system will fail if it doesn't fit an institution's procurement rules, governance capacity, or organisational culture. Identity systems have to be designed for the bureaucracy they live inside.
Human
Records that can absorb change: Names change. Gender markers change. Disability status changes. Households change. People move, age, and lose or gain legal status. Infrastructure moves slowly, by design: procurement cycles, legacy schemas, institutional caution. People don't. A system built for a life that holds still will quietly exclude the person it was meant to serve.
Digital infrastructure has a slow temporality: legacy schemas, technical debt, procurement cycles. Institutions have a medium temporality: policy cycles, budget cycles, leadership turnover. People have a fast temporality: names change, genders change, disability status changes, lives change. My work sits in that gap: building socio-technical systems robust enough to hold all three.
What Collaborators Say
Marieke Bierhoff, Director HIV @ Global Medical Affairs: I worked together with Katherine when we were both living and working in a very rural area in Thailand. Katherine has amazing skills as engineer but also has the ability to make very complex aspects of this understandable for everyone. She talks professional but understandable which is a rare gift. She has a creative and growth mindset and anybody would be lucky to have her in their team.
About
I'm Katherine Wyers, a software engineer (Python, Java Spring Boot, React, PHP) who recently completed a PhD in Digital Public Infrastructure at the University of Oslo, and I'm now an affiliated researcher at the University of Edinburgh. I specialise in identity data, meaning how it's collected, linked, deduplicated, and governed across health, education, and civil registration systems.
That specialisation started in the field, not in the literature. For a decade before the PhD, I worked hands-on in global health and education technology, as a founder, as a Senior Software Engineer building patient and student data systems with NGOs in Thailand and Myanmar, and as a consultant stabilising a provinical education management system rollout in South Africa. Across the projects, the hard problem was always the same shape: matching and maintaining identity records across fragmented systems, often offline, often for people whose lives didn't fit the database schema. My PhD took that field problem and studied it at scale, examining how the recategorisation of gender in Indian government identity systems, following the 2014 NALSA judgement, reveals where infrastructure fails the people it's meant to serve.
That combination is what I bring to consulting now. Right now, I'm applying it directly as identity data specialist on the early design of a DHIS2-based identity tracking system for a large international NGO, alongside production contributions to the DHIS2 codebase. Field-level engineering experience with real identity data systems, and research-level understanding of why identity infrastructure succeeds or fails. I'm building my practice for organisations that want technical depth on identity data specifically, not a generalist.
I'm currently based in Edinburgh, working globally, and taking on new consulting engagements in digital public infrastructure, identity systems, and humanitarian data platforms.
Publications and Speaking
My PhD research examined how identity infrastructure can exclude the populations it's meant to serve, the foundation for how I approach systems design today.
Selected Publications
- Making Personal Data Personal: The Emergence of Enactive Data in Organizations, Book Chapter in “Technology, Management, and Design for Social Justice”, Springer Books, 2026.
- Information Infrastructural Violence: A critical, intersectional study of gender recategorization in government information infrastructures, PhD Thesis, University of Oslo, 2025.
- Identity Categories as Boundary Objects: Conflict and Resistance in Collaboration, Best Paper Award, Scandinavian Conference on Information Systems, 2025.
- Torqued by an Identity Recategorization: The twisting consequences of rewriting identity categories in information infrastructure. Work in progress.
- Managing Discretion at the Street Level of Information Systems Use, AIS SIG Global Development, 2024.
- Health ICTs and Transgender Health Equity: A Research Agenda, Information Technology for Development Journal, 2024.
- Identity Platforms and Anti-LGBTQ+ Legislation: Implications for Safeguarding Personal Data, Joint IFIP Conference, 2023.
- Leaving No-one Behind? A Research Agenda for Queer Issues in ICT4D, IFIP Conference, 2022.
Speaking
- Mar 11th 2026: Digital Public Infrastructures and the State Recategorization of Human Identity Categories: Association of Information Systems UK.
- Jan 27th 2026: Driving Organizational Performance using an LGBTQ+ Inclusive HR Strategy: Global Diversity and Equality Management, Human Resources Management Masters Course, University of Edinburgh, Scotland.
- Nov 21st 2025: Trans Lives and the Slow Violence of Digital Infrastructure: Trans Research Association Ireland Annual Symposium, Belfast, Northern Ireland.
- Nov 18th 2025: Information Infrastructural Violence: Public Defence of Doctoral Thesis, University of Oslo, Norway.
- Oct 7th 2025: Are Your Digital Records Erasing Identities? Rethinking Gender Categories in Electronic Health Records: Digital Health, Masters-level course, University of Oslo, Norway.
- Sept 23rd 2025: Digital Infrastructures and Trans* Equity: Challenges and Pathways: ICT and Global Inequalities Masters-level course, University of Oslo, Norway.
From the Blog
- August 12, 2026: Git Commands Cheat Sheet: The Commands You’ll Use Every Day
- August 3, 2026: DHIS2 Docker Local Setup Script
- August 3, 2026: When Digital Identity Mismatches Cause Systemic Harm: Infrastructure Challenges and Legal Gaps
- June 30, 2026: 15 Coding Patterns for Solving Technical Interview Coding Questions
- February 6, 2026: How to improve your data-collection and IT systems to make your organization more efficient
Frequently Asked Questions
- What is a Digital Public Good? What's the DPG Alliance?
- A Digital Public Good is open-source software, data, or AI models designed to help meet the UN Sustainable Development Goals. DHIS2, OpenCRVS, and MOSIP are examples. The DPG Alliance is the body that certifies and supports this ecosystem. I work with implementers of these systems specifically, on the identity data layer.
- How is this different from general data consulting?
- Most data consultants treat identity as a technical problem. Deduplication logic and matching algorithms. I treat it as three problems stacked on top of each other: the technical one, the institutional one (procurement, governance, capacity), and the human one (people's identities change, and the system has to hold that). Missing any one of the three is usually why identity projects stall.
- Do you work with any specific platforms?
- I'm a DHIS2 contributor and I know its architecture closely, and I'm currently applying that as identity data specialist on a DHIS2-based identity tracking implementation. I have worked as a technical consultant on an OpenEMIS provincial implementation, and I follow MOSIP and OpenCRVS closely, since they're core to my specialisation. If it's a different platform, I'd still want to hear about the problems, because many of the identity linkage issues are platform-agnostic.
- Is my projects a good fit, or is this out of scope?
- Book a call and I'll tell you honestly. I'd rather point you toward the right person than take on something outside what I actually know well.
Contact
Currently accepting new consulting engagements, advisory work, and implementation roles