calebed.meBUILD // CASE STUDY

ACTIVE · HTML · Serverless

KJ ResearchOS

Qualitative research platforms either overwhelm you with spreadsheets or automate synthesis with AI. ResearchOS runs locally, organising raw data into an interactive board that preserves the necessary friction of deep analysis.

KJ ResearchOS view 1
KJ ResearchOS view 2

Evidence

Excerpts Analysed

589

Codes Ingested

33

Personal Time Saved

240min

Reflection

I built this workspace because I was frustrated by the tools available for my own academic research. Standard qualitative analysis tools either buried me in flat, overwhelming spreadsheets or asked me to outsource the synthesis to AI models. Handing my raw data over to an algorithm felt like skipping the most critical phase of my own learning; I wanted an environment that made cross-referencing instant while still forcing me to actively wrestle with the human narratives. The result is ResearchOS, a local-first digital workspace that organises transcripts and codebooks into a responsive, interactive board. It lets a researcher pivot across cultural contexts and participant voices in real time without losing their personal connection to the data. I engineered the application to run entirely inside the browser using client-side CSV parsing. This choice keeps the interface exceptionally fast and ensures that sensitive participant interviews remain strictly private. To prevent visual fatigue, I designed a context pane that surfaces codebook definitions only when they are relevant to the specific text block in view. My ongoing focus is figuring out how to transition this from a bespoke utility into a flexible, open architecture where other scholars can map their own conceptual parameters dynamically.