Knowledge Base
How can we help? 👋
All Categories
Platform Essentials
Using Lumi: writing effective prompts

Using Lumi: writing effective prompts

Write Lumi prompts that get accurate, useful answers — with example libraries for portfolio analysis, data quality, decarbonisation, reporting, and DDQs.

Last updated on April 25, 2026

Purpose of this article

Lumi is Scaler's built-in AI assistant for natural-language queries across your portfolio data and Scaler's documentation. This article explains how to write effective prompts, refine outputs, and use Lumi for the workflows where it adds the most value.

For the broader picture of how AI works in Scaler, see AI in Scaler: overview.

✨

Lumi AI is part of the Scale Plan. Every user receives 15 prompts/month free; more requires the Scale Plan — contact your account manager. Some organisations have AI features disabled. Access via the floating icon, the homepage, or inline throughout the platform.

Understanding Lumi's access scope

Lumi operates entirely within Scaler — it analyses and summarises the data you already have access to in the platform.

  • It does not access external data sources or third-party content
  • Its responses are scoped to the data and metrics your user permissions allow
  • For missing or incomplete data, Lumi highlights gaps and trends rather than inventing values

For the full security and architecture story, see AI in Scaler: overview.

Writing effective prompts

Be specific

The more context you include, the better Lumi can tailor its response. Where possible, include:

  • The asset, portfolio, or scope you mean
  • A time period (e.g. "FY2024" or "the last quarter")
  • The metric type (e.g. total_ghg_emissions, energy_consumption)

Examples:

  • "Summarise Portfolio Alpha's FY2024 energy use by asset."
  • "Compare Scope 1 and Scope 2 emissions for Building A over the last three years."

Lumi automatically interprets data available in Scaler, but performs best when you specify the exact scope of your query.

Request a specific output format

You can guide Lumi to present results in a preferred layout. Examples:

  • "List these findings as bullet points."
  • "Show this in a 3-column table — asset, total GHG, change versus last year."
  • "Summarise this in a short paragraph for a board report."

Structured output requests help Lumi match your reporting and communication standards.

Refine and follow up

Lumi retains conversation context within a session, so you can build on previous answers. Examples:

  • "Break this down by building."
  • "What's driving this trend?"
  • "Compare these results against targets."
  • "Summarise just the top five contributors."

Tip: If a first answer is not useful, refine the prompt with more specific scope, a clearer time period, or an explicit output format. You'll usually get a much sharper second answer.

Example prompt library, by use case

Portfolio analysis and risk

  • "Analyse the top 10 worst performing assets by energy intensity. Show the result as a table and a chart."
  • "What risks — environmental and financial — is my portfolio facing, and how can I best mitigate them?"
  • "Summarise the overall sustainability performance of my portfolio."
  • "Which assets in my portfolio have the highest transition risk?"

Data quality and gap analysis

  • "Identify missing data points across assets for the current reporting year."
  • "Which meters in my portfolio have the largest data coverage gap?"
  • "Show assets where like-for-like values are null despite 100% data coverage."
  • "Flag any assets with implausible energy-to-GHG ratios."

Decarbonisation and CapEx planning

  • "Which retrofit measures would have the biggest impact on this portfolio's emissions?"
  • "Compare planned CapEx against the CRREM 1.5°C pathway for this portfolio."
  • "Which five assets are most pressing for action over the next three years?"
  • "What's the most cost-effective measure to bring this asset back onto its pathway?"

Reporting and framework readiness

  • "Summarise how our disclosures align with CSRD, SFDR, and GRESB indicators."
  • "What information do I still need to deliver to lift our GRESB score?"
  • "Which green certifications are expiring in the next 12 months?"
  • "Compare energy intensity across portfolios and flag the underperformers."

Investor / DDQ responses

  • "Draft a response to: 'What is the fund's total energy consumption and carbon footprint?'"
  • "Describe our approach to net-zero target setting, using portfolio data."
  • "What percentage of assets have undergone energy audits in the past three years?"
  • "How do we monitor and manage physical climate risks across the portfolio?"

Platform help and methodology

  • "How does Scaler calculate my GHG emissions and determine which scopes they belong to?"
  • "Explain how data coverage is calculated."
  • "What's the difference between location-based and market-based emission factors in Scaler?"
  • "Where do I configure portfolio-level emission factors?"

Where to find Lumi

Lumi is available from the Home page in the left-hand navigation toolbar. The default landing is Welcome, with a Lumi prompt entry and category tabs — Analyze, Report, Compare, Improve — each with suggested prompts.

You can also open Lumi from the flower icon in the corner of the sidebar, visible from anywhere in the platform. The icon position is configurable — drag it between the bottom-left and bottom-right corners.

Notion image
Notion image

Tips for consistent results

  • Make sure portfolio and asset data are up-to-date in the Data Collection Portal before querying Lumi for portfolio-wide analysis
  • For multi-year trends, confirm the reporting period selected in the Analytics Portal matches the timeframe in your prompt
  • Use short, direct phrasing — Lumi interprets clear task-based questions most accurately
  • Treat AI-generated suggestions, forecasts, and DDQ answers as drafts to verify, not finished outputs

Where to go next

Did this answer your question?
😞
😐
🤩