AHRC · July 2025–present
A technology transformation at AHRC
As IT Business Systems Manager and Solution Architect at the Afghanistan Human Rights Center, I lead a five-person team and own decisions about infrastructure and systems architecture. The work began with fragmented tools and manual processes. I translated the board’s needs into technical specifications and a phased delivery plan, then led the work to connect the organization’s infrastructure, data, and workflows.
This is an account of professional practice. It describes the systems we built and the research questions that experience has raised.
Establishing the foundations
I took responsibility for the domain, website, Cloudflare, Salesforce, and collaboration tools. I evaluated vendors against their capabilities, integration requirements, and reliability. I also migrated organizational email and files from Microsoft to Google Workspace so staff could share documents and collaborate online more easily.
These decisions shaped what we could build next. Shared access, clearer ownership, and connected records were necessary foundations for useful automation.
Connecting the systems
I built and deployed an integration between Every.org webhooks and the Salesforce API to create donor and gift records automatically for online donations. The process had relied on checks and manual entry into spreadsheets; the integration established a connected path for online transactions.
This kind of work is where my business systems background and technical practice meet: define the workflow, map the data, implement the integration, and check that the records support the business process.
Adding AI with review and approval
Our incident workflow uses the Claude API to help turn source material into structured records for staff review. For reporting, retrieval-augmented generation brings relevant incident records into context so Claude can produce cited drafts. Staff can review and edit the draft before approval and publication.
Weekly board-report preparation previously involved pulling records, reviewing the week’s incidents, and assembling material for the board. Based on my operational experience, the retrieval and draft-preparation work fell from at least four hours a week to near zero. Human review and approval remain part of the workflow.
That time saving is a practitioner estimate. It is not a controlled measurement of AI’s effect separately from the infrastructure and workflow improvements.
What this experience makes me want to study
The transformation crossed several layers: infrastructure, data, integration, AI, and the way people review work. It makes me want to examine which layer is responsible for an improvement, what evidence an automated system should provide, and how the whole process recovers when something goes wrong.
Those questions are the starting point for my research on execution evidence and recovery decisions.
Earlier experience
IT Business Systems Analyst. Coordinated seven teams on an authentication migration covering 135,000 eligible accounts, with responsibility for acceptance testing and go/no-go decisions, including a rollback. My work also included SQL, data pipelines, API troubleshooting, Python monitoring, and an internal knowledge site that reduced routine questions from business partners by about 60%.
IT Business Systems Analyst. Led a six-person software team through the development lifecycle, translated workflows into user stories, and built Tableau dashboards. I used Python to automate cleaning, validation, deduplication, and field mapping for Salesforce imports, reducing weekly preparation from four hours to a few minutes.
My professional portfolio provides a broader view of my experience and technical work.