You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Enable users to select a user persona when generating a dashboard. This selection will customise the summary text and insights to better match the user’s needs, skills, and domain expertise. This is a core value proposition for this prototype and we want to test the hypothesis that LLM-driven outputs can be tailored based on the user, is valuable, and improves accessibility of data insights.
Users can choose from predefined personas or define their own.
Predefined Personas
Each persona includes a short title (for the dropdown) and a detailed description (available via an info button or inline display).
Conservation Manager, Biodiversity & Threat Monitoring Oversees Key Biodiversity Areas, with basic GIS skills. Needs to identify threats (e.g., illegal logging, habitat degradation) and monitor ecosystem health to plan conservation efforts.
Program Manager, Nature-Based Solutions & Restoration Works on agroforestry and land restoration. Uses QGIS for mapping. Needs to track tree cover gain, carbon sequestration, and prioritise interventions based on risks like erosion or forest loss.
Investment Analyst, Impact & Risk Monitoring Supports reforestation and agroforestry investments. Limited GIS skills, relies on dashboards. Needs geospatial insights to monitor portfolio performance, assess risks, and ensure alignment with net-zero goals.
Sustainability Manager, Supply Chain Compliance Ensures agricultural supply chains meet deforestation-free commitments. Limited GIS skills, uses web-based tools. Needs to monitor land conversion risks and support sustainable sourcing decisions.
Advocacy Manager, Land Rights & Policy Works on Indigenous land rights advocacy. Basic GIS skills for visualisation. Needs data to track land use changes, push for policy reforms, and empower local communities to monitor territories.
Environmental Journalist, Corporate Accountability Covers environmental and corporate accountability issues. Basic GIS skills to interpret data visually but not analyze. Needs accessible data to track corporate EUDR compliance and produce data-driven reports.
Custom Persona Option
If none of the predefined personas fit, users can enter their own details:
Who are you? e.g. "I am a [role] responsible for [responsibilities]. I have [GIS skill level], meaning I can [GIS capabilities]."
What do you need? e.g. "I need to [geospatial need], such as [specific tasks or challenges]. This helps me [desired outcomes]."
How insights are used? (resource allocation, impact assessment, advocacy... etc)
Scope
Define the functional boundaries of this feature.
In
A simple dropdown with predefined persona names.
An info button to show a full persona description (could be simple text, or Zeno generates a full description. the latter could be faked)
Ability to define a custom persona through a structured input form.
Outputs (e.g. portfolio report, text summaries for data visualisations) adapt to the selected persona.
Out
Persona customisation (e.g. learning from past conversations).
Direct integration with external user profiles (no upload etc)
User Stories & Acceptance Criteria
Selecting a Persona
As a user setting up a dashboard, I want to select a persona from a predefined list, So that the insights are tailored to my needs, skills, and domain expertise.
Acceptance Criteria:
The dropdown provides a list of personas.
An info button shows the full description.
The system applies the persona selection to adjust summaries and insights.
Understanding Personas
As a user unfamiliar with the personas, I want to see a clear description of each persona before selecting one, So that I can choose the most relevant option.
Acceptance Criteria:
Clicking an info button reveals a detailed description.
The description explains the persona’s skills, needs, and use cases (static text, or Zeno driven)
Custom Persona Definition
As a user who doesn’t fit into the predefined personas, I want to define my own persona, So that my dashboard insights are more relevant to my specific role and needs.
Acceptance Criteria:
Users can enter custom responses following a structured template.
Custom personas adjust outputs in the same way as predefined ones.
Persona-Based Output Generation
As a user who selected a persona, I want to receive insights and summaries tailored to my persona’s expertise, So that the outputs are understandable and actionable for me.
Acceptance Criteria:
Reports and data viz summaries use persona-specific language and framing.
Reports and data viz summary language aligns with the user’s level of technical/domain expertise.
Reports and data viz summaries incorporate relevant KBA data (e.g. species, land management etc)
Risks & Questions
Accessibility
How do we communicate that persona selection only affects the language of outputs, not the underlying data?
Usability
Should we allow users to edit predefined personas (e.g., adjust skill levels)?
Should the dropdown pre-select a default persona, or require explicit selection?
Feasibility
How dynamic should the persona-based text adaptation be?
Can we provide example summaries for each persona to demonstrate differences?
Solution Hints
Use a structured dropdown with info buttons for clarity.
Offer a guided process or conversation for defining a custom persona.
Ensure real-time preview of how persona selection affects outputs.
Future Work
Enhanced Personalisation
AI-driven persona recommendations based on ongoing conversations.
Dynamic adaptation of visualisations (not just text).
User Experience Improvements
Ability to save custom personas for future use.
More granular customisation (e.g. adjusting GIS skill level within a persona).
The text was updated successfully, but these errors were encountered:
Goal
Enable users to select a user persona when generating a dashboard. This selection will customise the summary text and insights to better match the user’s needs, skills, and domain expertise. This is a core value proposition for this prototype and we want to test the hypothesis that LLM-driven outputs can be tailored based on the user, is valuable, and improves accessibility of data insights.
streamlit example
Users can choose from predefined personas or define their own.
Predefined Personas
Each persona includes a short title (for the dropdown) and a detailed description (available via an info button or inline display).
Conservation Manager, Biodiversity & Threat Monitoring
Oversees Key Biodiversity Areas, with basic GIS skills. Needs to identify threats (e.g., illegal logging, habitat degradation) and monitor ecosystem health to plan conservation efforts.
Program Manager, Nature-Based Solutions & Restoration
Works on agroforestry and land restoration. Uses QGIS for mapping. Needs to track tree cover gain, carbon sequestration, and prioritise interventions based on risks like erosion or forest loss.
Investment Analyst, Impact & Risk Monitoring
Supports reforestation and agroforestry investments. Limited GIS skills, relies on dashboards. Needs geospatial insights to monitor portfolio performance, assess risks, and ensure alignment with net-zero goals.
Sustainability Manager, Supply Chain Compliance
Ensures agricultural supply chains meet deforestation-free commitments. Limited GIS skills, uses web-based tools. Needs to monitor land conversion risks and support sustainable sourcing decisions.
Advocacy Manager, Land Rights & Policy
Works on Indigenous land rights advocacy. Basic GIS skills for visualisation. Needs data to track land use changes, push for policy reforms, and empower local communities to monitor territories.
Environmental Journalist, Corporate Accountability
Covers environmental and corporate accountability issues. Basic GIS skills to interpret data visually but not analyze. Needs accessible data to track corporate EUDR compliance and produce data-driven reports.
Custom Persona Option
If none of the predefined personas fit, users can enter their own details:
Scope
Define the functional boundaries of this feature.
In
Out
User Stories & Acceptance Criteria
Selecting a Persona
As a user setting up a dashboard,
I want to select a persona from a predefined list,
So that the insights are tailored to my needs, skills, and domain expertise.
Acceptance Criteria:
Understanding Personas
As a user unfamiliar with the personas,
I want to see a clear description of each persona before selecting one,
So that I can choose the most relevant option.
Acceptance Criteria:
Custom Persona Definition
As a user who doesn’t fit into the predefined personas,
I want to define my own persona,
So that my dashboard insights are more relevant to my specific role and needs.
Acceptance Criteria:
Persona-Based Output Generation
As a user who selected a persona,
I want to receive insights and summaries tailored to my persona’s expertise,
So that the outputs are understandable and actionable for me.
Acceptance Criteria:
Risks & Questions
Accessibility
Usability
Feasibility
Solution Hints
Future Work
Enhanced Personalisation
User Experience Improvements
The text was updated successfully, but these errors were encountered: