Back to all workSources selected4 Topic filters12 Review statusOpen
Custom applications
Customer research dashboard
Turning scattered online discussions into research people can review.
Useful conversations were spread across public communities and difficult to compare. We built a research workspace that gathers selected discussions, gives them a consistent structure and keeps the original evidence available for human judgement.
Research overview
Discussion library
Discussion themePrimary topicUrgency
Managing seasonal stockInventory7 / 10
Explaining delivery choicesCheckout6 / 10
Keeping product data tidyOperations8 / 10
Comparing support toolsCustomer care5 / 10
Fictional topics and scores
The requirement
A folder of links was not enough to reveal useful patterns.
Once research covered more than a handful of conversations, the material became hard to revisit and compare. The team needed a shared structure for exploration while preserving the source context that makes research credible.
- Bring selected public discussions into one workspace
- Give varied posts a consistent research structure
- Keep the original evidence available for human review
- Help researchers filter by topic, source and urgency
- Give administrators control over users and search terms
The research journey
Collect, structure, review and investigate.
Automation helps organise the material. Researchers remain responsible for deciding what it means and what deserves closer attention.
- 01
Collect
A researcher chooses a supported source and starts a focused collection run using approved search terms.
- 02
Structure
The application organises each distinct discussion into the same set of research fields.
- 03
Review
Researchers filter the material, compare themes and inspect the source evidence behind each record.
- 04
Investigate
Promising patterns can be explored further without treating an automated score as a final conclusion.
What we built
A thoughtful research product, not an answer machine.
The dashboard helps people organise and explore discussions. It deliberately keeps analysis connected to the evidence it came from.
Guided collection
Researchers can collect relevant public discussions from selected sources and follow progress as the run completes.
Consistent structure
Each discussion is organised around pain points, topics, urgency, business context, summary and supporting evidence.
Flexible exploration
Filters help researchers move between sources, dates, topics and urgency without losing the underlying conversation.
Human judgement
Source context and supporting excerpts stay visible so the researcher can question, compare and interpret the analysis.
Research record
ReviewKeeping the evidence attached
A structured summary helps comparison, while the source discussion remains available for the researcher to assess in context.
- Primary topicInventory
- Reported problemSeasonal stock planning
- Urgency7 / 10, triage only
- Next actionReview source context
Fictional example
The outcome
Scattered conversations became structured material for human review.
The application gives researchers a clearer route from collection to investigation while keeping automated analysis in its proper supporting role.
Selected sources
Relevant public discussions in one workspace
Shared structure
Consistent fields for comparison
Source evidence
Original context available for review
Human judgement
Scores used for triage, not proof
Technology
The technology behind the work.
Next.js- TypeScript
- Prisma
- OpenAI API
- Secure access
- Tailwind CSS
About this project
- Sector
- Ecommerce research
- Services
- Custom application development, research workflow and AI-assisted analysis
- Project type
- Internal research platform
Need a better way to explore business research?
Tell us where the evidence lives and how your team needs to review it. We will help shape a practical workflow around both.