How vendors are chosen & classified?

SYSTEM:
You are an expert analyst for an energy technology directory. Return valid JSON.

USER:
I will provide a list of URLs from {{ vendor_name }}. Select a maximum of 10 URLs most likely to be core product or service pages. Ignore blogs, news, careers, and generic pages.

{{ vendor_context }}

Return a JSON object matching this schema:
{
  "selected_urls": ["string"]
}

URLs:
{{ url_list }}

SYSTEM:
You are an expert analyst for a European energy technology directory. Use factual, professional language without marketing superlatives. If the homepage represents the core product or platform, you may extract it using the homepage URL. Return valid JSON with an 'offerings' key.

USER:
Analyze the following webpage contents for {{ vendor_name }} and extract their core packaged software products and platforms. CRITICAL: Select a MAXIMUM of 5 products that are packaged software solutions, platforms, or defined technology offerings. You may select fewer than 5 if there are not enough distinct packaged products.

PRODUCT URL RULE:
If the vendor's website is simple (e.g. a single-product company or startup) and the homepage itself represents the primary product or platform, you are explicitly allowed and encouraged to include the homepage URL (marked with HOMEPAGE) as the product's URL.

Return a JSON object matching this schema:
{
  "offerings": [
    {
      "name": "string",
      "description": "string",
      "url": "string"
    }
  ]
}

Contents:
{{ pages_contents }}

SYSTEM:
You are an expert analyst for a European energy technology directory. Your task is to identify business cases, reference projects and success stories from URL lists. Return valid JSON.

USER:
I will provide a list of URLs from the website of {{ vendor_name }}, an energy technology company.

Context about the company's technical domains and offerings:
{{ tech_context }}
{{ vendor_context }}

Select a maximum of 10 URLs that are MOST LIKELY to be business cases, reference projects, success stories, case studies, or project announcements related to these technologies.

PRIORITIZE by:
1. PROJECT SCALE (MW) — URLs suggesting large MW-scale projects or containing keywords like 'utility-scale', 'grid-scale', 'transmission', 'giga', 'mega' rank higher
2. RECENCY — URLs suggesting recent dates (2024-2026) rank higher
3. PARTNERSHIPS — URLs mentioning partnerships, collaborations, or joint ventures rank higher

IGNORE: product pages, career pages, legal/privacy pages, API docs, sitemaps, and cookie/privacy policy pages.

Return a JSON object:
{
  "selected_urls": [
    {
      "url": "string",
      "rationale": "string (brief reason why this URL was selected)"
    }
  ]
}

URLs:
{{ url_list }}

SYSTEM:
You are an expert analyst for a European energy technology directory. Extract business cases using factual, professional language without marketing superlatives. Return valid JSON with an 'articles' key.

USER:
Analyze the following webpage contents for {{ vendor_name }} and extract their business cases, reference projects, and success stories.

Context about the company's technical domains and offerings:
{{ tech_context }}
{{ vendor_context }}

Select a MAXIMUM of 5 articles that represent the most impactful business cases.

PRIORITIZE by:
1. PROJECT SCALE (MW) — Larger MW capacity projects rank higher. You MUST locate and extract the capacity in MW where available.
2. RECENCY — More recent articles rank higher
3. PARTNERSHIPS — Articles mentioning partnerships or collaborations rank higher
4. CONCRETE RESULTS — Articles with specific numbers, outcomes, or metrics rank higher

Return a JSON object:
{
  "articles": [
    {
      "title": "string",
      "date": "YYYY-MM-DD or null if unknown",
      "url": "string",
      "capacity_mw": "number (float) or null if unknown/not specified",
      "excerpt": "string (2-3 sentence summary focusing on the business case)"
    }
  ]
}

Sort the final list of articles by `capacity_mw` in descending order (with nulls last) so that the largest scale projects are at the top.

Contents:
{{ pages_contents }}

SYSTEM:
You are an expert energy tech analyst. Your task is to:
1. Classify a vendor into exactly ONE category (technology) and between 1 and 3 subcategories (sub_technologies) linked STRICTLY to that category.
2. Estimate the probability distribution (category_probabilities) across ALL 3 categories in the taxonomy based on how much the vendor's products, services, and reference articles map to each category.
3. Assign an independent score between 0 and 100 for EVERY single subcategory across ALL categories in the entire taxonomy in "subcategory_scores". 0 is very unlikely, 100 is highly likely / exact match. These are independent scores, so they do NOT sum to 100.
4. Generate a short positioning snapshot ("llm_summary") in English, 120–200 characters max.
   Rules for llm_summary:
   - Do NOT start with the vendor name (it's redundant).
   - Keep the summary between 120 and 200 characters.
   - Use exactly 3 bold keywords wrapped in **double asterisks**: one for the market SEGMENT, one for the key DIFFERENTIATOR, and one for the TARGET customer type.
   - Mention European market specifics (e.g. EPEX Spot, aFRR/mFRR, Redispatch 2.0, Spain's NTS 2.1) ONLY when they are a genuine differentiator for this vendor.
   - Tone: strictly factual and technical. Written for a European energy professional audience. NO superlatives, NO marketing adjectives (e.g. "ultra-accurate", "cutting-edge", "best-in-class", "world-class", "unparalleled", "game-changing"). Describe WHAT the vendor does and HOW, not how well.
   - Good example: "**Solar+wind production forecasting** combining NWP models and satellite data — reduces **imbalance exposure** for **TSOs and traders**."
   - Bad example: "**Ultra-accurate forecasting** with best-in-class AI — delivers unmatched predictions for **energy leaders**."
    
{{ taxonomy_prompt }}

CLASSIFICATION & PROBABILITY RULES:
1. You MUST select exactly ONE category from the TAXONOMY DEFINITION for "technology".
2. You MUST select between 1 and 3 subcategories for "sub_technologies".
3. All selected "sub_technologies" MUST belong to the chosen "technology" category.
4. You MUST rate the probability distribution for ALL 3 categories in "category_probabilities".
5. Every category name in the taxonomy must be a key in "category_probabilities" (even if the probability is 0).
6. Each probability must be an integer (no decimals).
7. The sum of all probabilities across the 3 categories must equal exactly 100.
8. You MUST rate every single subcategory across ALL categories in the taxonomy on a 0-100 scale in "subcategory_scores". The keys must be EXACTLY the names of all subcategories in the entire taxonomy.
9. You MUST generate a positioning snapshot between 120 and 200 characters containing exactly 3 bold keywords in **double asterisks** in "llm_summary".
10. You must output your decision strictly as a JSON object with keys: "technology", "sub_technologies", "category_probabilities", "subcategory_scores", and "llm_summary".

Example output format:
{
  "technology": "Markets & orchestration",
  "sub_technologies": ["Energy trading & execution", "Flexibility & balancing services"],
  "category_probabilities": {
    "Markets & orchestration": 70,
    "Analytics & Forecasts": 20,
    "Monitoring & control": 10
  },
  "subcategory_scores": {
    "Multi-asset VPPs": 90,
    "Industrial demand response": 85,
    "Grid-scale BESS optimization": 10,
    "Utility-scale renewables trading": 0,
    "Energy market intelligence": 0,
    "Renewable power forecasting": 0,
    "Consumption & load forecasting": 0,
    "Price forecasting": 0,
    "Investment & asset valuation": 0,
    "Battery analytics & diagnostics": 0,
    "Weather & resource assessment": 0,
    "Energy efficiency analytics": 0,
    "PPA & procurement analytics": 0,
    "On-site EMS / edge controller": 20,
    "SCADA": 5,
    "Power plant controller (PPC)": 0,
    "Asset performance management": 10,
    "O&M / CMMS": 0,
    "Building & C&I energy management": 0
  },
  "llm_summary": "**Renewable orchestration** via cloud VPP — **BSP scheduling** for ancillary markets (aFRR/mFRR), targeting **aggregators**."
}

USER:
Please classify the following vendor '{{ vendor_name }}' based on their profile, products, and articles content:

{{ vendor_context }}

categories:
  - name: Markets & orchestration
    description: Software that connects MW-scale energy assets to wholesale and balancing markets — covering algorithmic trade execution, VPP aggregation, industrial demand response, and grid-scale BESS optimization.
    relevant_for:
      - Real-time fleet dispatch and market bidding
      - Automated or algorithmic trading on spot/intraday markets
      - VPP aggregation of diverse MW-scale distributed assets
      - Industrial demand-side flexibility and load curtailment
      - Grid-scale BESS multi-market revenue stacking
    belongs_elsewhere:
      - Behind-the-meter-only energy efficiency (no market participation)
      - Pure forecasting without trade execution
    see_also: Analytics & Forecasts (for forecasting inputs used by trading)
    subcategories:
      - name: Algorithmic Power Trading & ETRM
        description: Software for automated market access, algorithmic trading, and Energy Trade & Risk Management (ETRM). Manages spot/intraday bidding, position tracking, and imbalance risk across power portfolios without direct physical asset control.
        relevant_for:
          - Algorithmic and automated spot/intraday trading
          - Direct market access (e.g., EPEX, Nord Pool)
          - Multi-market position and P&L tracking
          - Energy Trade & Risk Management (ETRM)
        belongs_elsewhere:
          - Physical dispatch of distributed assets → Utility & C&I VPP Aggregation
          - Pure price forecasting without execution → Analytics & Forecasts > Price forecasting
      - name: Grid-scale BESS & Hybrid plants
        description: Dedicated software for maximizing revenue of utility-scale battery energy storage systems (BESS) and co-located hybrid renewable plants. Focuses on AI-driven multi-market dispatch while managing site-level constraints and battery degradation.
        relevant_for:
          - Grid-scale BESS revenue stacking (wholesale + ancillary)
          - Co-located hybrid asset (Renewables + BESS) trading
          - AI/ML-driven dispatch and degradation-aware bidding
        belongs_elsewhere:
          - Aggregation of remote, distributed small-scale batteries → Utility & C&I VPP Aggregation
          - Battery health analytics and diagnostics → Analytics > Battery analytics & diagnostics
      - name: Demand Response & Load Flexibility
        description: Platforms tailored to monetize consumption flexibility from industrial, commercial, and municipal loads. Focuses on safe load curtailment, process shifting, and bidding into ancillary services without disrupting core operations.
        relevant_for:
          - Industrial load curtailment and process shifting
          - Commercial building flexibility (HVAC, refrigeration)
          - DSO flexibility procurement and congestion management
        belongs_elsewhere:
          - Energy efficiency analytics (no market participation) → Analytics > Energy efficiency analytics
          - High-volume residential mass-market aggregation → Residential & Fleet Aggregation
      - name: Utility & C&I VPP Aggregation
        description: Cloud platforms that aggregate commercial-scale and utility-scale distributed generation and storage to act as a single market participant. Focuses on physical remote dispatch and portfolio optimization for balancing markets.
        relevant_for:
          - Aggregation of distributed generation and commercial storage
          - Remote dispatch of diverse distributed energy pools
          - Cross-asset portfolio optimization for balancing markets
        belongs_elsewhere:
          - Optimization of single utility-scale sites → Grid-scale BESS & Hybrid plants
          - Aggregation of mass-market residential assets and EVs → Residential & Fleet Aggregation
      - name: Residential & Fleet Aggregation
        description: High-volume cloud orchestration of mass-market, small-scale assets (EV smart charging, home batteries, residential solar, heat pumps), primarily serving energy retailers and fleet operators.
        relevant_for:
          - EV smart charging and fleet aggregation
          - Residential battery and heat pump orchestration
          - Retailer-led virtual power plants (VPP)
        belongs_elsewhere:
          - Commercial & industrial distributed asset aggregation → Utility & C&I VPP Aggregation

  - name: Analytics & Forecasts
    description: Platforms and services that produce forecasts, analytics, market intelligence, and investment-grade data — covering weather-to-wire power predictions, energy market data, asset health diagnostics, energy efficiency analytics, and business-case modelling.
    relevant_for:
      - Weather-driven power production predictions
      - Day-ahead, intraday, and long-term electricity price curves
      - Energy market data feeds, indices, and research reports
      - Battery state-of-health, degradation modelling, and safety scoring
      - Investment-case modelling and revenue stacking simulation
      - Industrial and building energy consumption analytics
      - Solar/wind resource assessment and bankable yield estimates
      - PPA valuation, certificate tracking, and procurement optimisation
    belongs_elsewhere:
      - Live trade execution or automated bidding
      - Real-time SCADA or plant control
      - VPP aggregation or demand response dispatch
    see_also: Markets & orchestration (when forecasts feed into automated trading)
    subcategories:
      - name: Energy market intelligence
        description: Data platforms, indices, and research services providing structured energy market information — covering real-time and historical commodity prices, market reports, forward curves, benchmarking indices, and macro-level energy statistics.
        relevant_for:
          - Energy market data feeds and price aggregation
          - Benchmarking indices (BESS revenue, PPA prices)
          - Research reports and market outlook publications
          - Global energy statistics and databases
        belongs_elsewhere:
          - Short-term price predictions → Price forecasting
          - Live trading execution → Markets & orchestration > Algorithmic Power Trading & ETRM
          - PPA-specific valuation and risk → PPA & procurement analytics
      - name: Renewable power forecasting
        description: Providers of primary, weather-driven power production forecasts delivered as standalone data products or APIs. These "pure-play" forecasting providers utilise proprietary meteorological data, NWP models, satellite imagery, and physics-constrained ML to predict generation for solar, wind, and hydro assets. This category is strictly for data providers selling forecasts, excluding software platforms (such as VPPs, trading systems, or EMS) that merely consume forecasts internally to optimize assets or execute market trades.
        relevant_for:
          - Generating primary power forecasts from raw meteorological data
          - Providing pure-play data feeds, APIs, or portals for power predictions
          - Operating multi-model NWP ensembles and physics-constrained ML
          - Selling raw forecasting data to TSOs, traders, or asset developers
        belongs_elsewhere:
          - Raw weather data and resource assessment → Weather & resource assessment
          - Electricity price predictions → Price forecasting
          - Live trading decisions based on forecasts → Markets & orchestration > Algorithmic Power Trading & ETRM
          - Internal forecasting used by VPPs, trading platforms, or EMS/SCADA → Markets & orchestration OR Monitoring & control
      - name: Consumption & load forecasting
        description: Providers of primary electricity demand and load predictions, delivered as standalone data products or APIs. This covers macro grid-level load forecasting, as well as building and EV charging demand predictions. This category is reserved exclusively for dedicated data providers and excludes downstream software platforms (such as VPPs or energy management systems) that utilize internal load forecasting solely to optimize assets or schedule flexibility.
        relevant_for:
          - Generating primary electricity demand predictions from raw data
          - Selling standalone load predictions via API or data portal
          - Operating grid-level macro demand models for TSOs/DSOs
        belongs_elsewhere:
          - Industrial energy efficiency recommendations → Energy efficiency analytics
          - Behind-the-meter load management/control → Monitoring & control > Building & C&I energy management
          - Internal load forecasting within an EMS or VPP → Markets & orchestration OR Monitoring & control
      - name: Price forecasting
        description: Dedicated market intelligence providers generating primary electricity price predictions—spanning from day-ahead clearing prices to 40-year fundamental curves. These forecasts are delivered as standalone data feeds or comprehensive reports. This category explicitly excludes algorithmic trading platforms, BESS optimizers, and VPPs that integrate price forecasting internally to execute automated market bids.
        relevant_for:
          - Generating primary electricity price predictions via market simulation
          - Selling price curves and fundamental scenarios as a standalone data product
          - Providing long-term (10–40 years) energy market price reports
          - Operating dedicated price forecasting APIs for traders
        belongs_elsewhere:
          - Historical price data feeds and indices → Energy market intelligence
          - Trading signals and bid optimization → Markets & orchestration > Algorithmic Power Trading & ETRM
          - BESS revenue projections combining price + dispatch → Investment & asset valuation
          - Internal price forecasting used to drive automated trading decisions → Markets & orchestration > Algorithmic Power Trading & ETRM
      - name: Investment & asset valuation
        description: Dedicated software products and SaaS platforms for clients to independently model, simulate, and value energy assets for investment, lending decisions, and bankability. This explicitly requires the company to provide a clear, client-facing product or tool (e.g., like Gridcog, Pexapark, Montel, Retwin, Clean Horizon, Aurora, Modo) to drive asset valuation, BESS revenue stacking simulations, IRR/NPV calculations, or PPA fair-value assessment.
        relevant_for:
          - Client-facing software tools and SaaS platforms for asset valuation
          - Simulation products used for lending and investment decisions / bankability
          - Software for BESS revenue stacking and storage valuation
          - Platforms for project feasibility, IRR/NPV modelling, and PPA risk quantification
        belongs_elsewhere:
          - Pure consulting, advisory, or custom reports (without a client-facing valuation software product) → Energy market intelligence
          - Raw weather data or bankable yield/resource estimates (e.g., Solargis, Meteodyn) → Weather & resource assessment
          - Production forecasting (e.g., ConWX) → Renewable power forecasting
          - Live BESS optimization, market execution, or congestion management (e.g., enspired, Skoon) → Markets & orchestration > Grid-scale BESS & Hybrid plants
          - PPA procurement and certificate management → PPA & procurement analytics
          - Real-time P&L tracking → Markets & orchestration > Algorithmic Power Trading & ETRM
      - name: Battery analytics & diagnostics
        description: Specialised analytics for battery health, safety, and lifetime — covering state-of-health estimation, degradation modelling, thermal runaway detection, digital twin simulation, warranty compliance monitoring, and commissioning testing (FAT/SAT).
        relevant_for:
          - Battery state-of-health (SoH) and state-of-charge (SoC) estimation
          - Degradation modelling and remaining useful life prediction
          - Thermal runaway detection and safety scoring
          - FAT/SAT commissioning and warranty compliance
        belongs_elsewhere:
          - BESS dispatch and revenue optimization → Markets & orchestration > Grid-scale BESS & Hybrid plants
          - General renewable APM (not battery-specific) → Monitoring & control > APM
          - BESS investment modelling → Investment & asset valuation
      - name: Predictive maintenance & AI analytics
        description: Dedicated AI solutions that ingest existing SCADA and operational data to provide unsupervised machine learning, anomaly detection, early failure prediction, and digital-twin-based root cause analysis for renewable assets like wind and solar.
        relevant_for:
          - AI-driven early failure prediction and anomaly detection
          - Unsupervised machine learning on SCADA data
          - Physics-informed AI and digital-twin root cause analysis
        belongs_elsewhere:
          - General fleet monitoring and KPI dashboards → Monitoring & control > Asset performance management
          - Any form of battery degradation, thermal safety, or BESS anomaly detection → Analytics > Battery analytics & diagnostics
          - Work order management and ticketing → Monitoring & control > O&M / CMMS
      - name: Weather & resource assessment
        description: Raw weather data services and bankable resource assessment for pre-construction and operational use — covering solar irradiance databases, wind resource maps, mesoscale/LES modelling, virtual met masts, hindcast datasets, and weather APIs.
        relevant_for:
          - Solar irradiance databases and satellite-derived data
          - Wind resource maps and virtual met mast simulations
          - Mesoscale and Large Eddy Simulation modelling
          - Hindcast datasets and long-term climate time series
          - Weather APIs for energy applications
        belongs_elsewhere:
          - Power production forecasting (MW output) → Renewable power forecasting
          - Energy price predictions → Price forecasting
          - On-site weather stations and sensors → out of scope
      - name: Energy efficiency analytics
        description: Analytics platforms for reducing energy consumption and carbon emissions in industrial facilities and commercial buildings — covering energy audits, consumption anomaly detection, process optimisation, ISO 50001 compliance, M&V (measurement and verification), and multi-site benchmarking.
        relevant_for:
          - Industrial energy intensity reduction and process optimisation
          - Building portfolio energy benchmarking and anomaly detection
          - ISO 50001 certification support and audit automation
          - Automated measurement and verification (IPMVP)
          - Multi-site energy consumption analytics and reporting
        belongs_elsewhere:
          - Active building control (HVAC, heat pumps) → Monitoring & control > Building & C&I energy management
          - Flexibility monetisation from industrial loads → Markets & orchestration > Demand Response & Load Flexibility
          - Energy bill management without analytics → out of scope
      - name: PPA & procurement analytics
        description: Tools for renewable energy procurement management — covering PPA valuation and benchmarking, energy attribute certificate tracking, 24/7 carbon-free energy matching, granular certificate management, and portfolio risk analytics for corporate buyers.
        relevant_for:
          - PPA benchmark pricing and fair-value assessment
          - Energy attribute certificate (EAC) tracking and matching
          - 24/7 carbon-free energy (CFE) procurement strategies
          - Corporate renewable procurement portfolio management
        belongs_elsewhere:
          - PPA route-to-market execution → Markets & orchestration > Algorithmic Power Trading & ETRM
          - Long-term price scenarios for PPA valuation → Price forecasting
          - ESG/carbon reporting → out of scope (adjacent)

  - name: Monitoring & control
    description: On-site and cloud-based systems for real-time monitoring, supervisory control, asset performance management, and maintenance of energy assets — spanning SCADA, power plant controllers, hardware EMS, and CMMS across utility-scale, C&I, and building applications.
    relevant_for:
      - SCADA, HMI, and real-time supervisory control
      - Power plant controllers for grid-code compliance
      - Fleet-wide asset performance dashboards and alarming
      - Maintenance scheduling, work orders, and ticketing
      - Edge controllers and on-site hardware gateways
      - Building and C&I behind-the-meter optimisation
      - Distribution grid DER management and congestion
      - Microgrid islanding and hybrid plant orchestration
    belongs_elsewhere:
      - Market-side portfolio optimisation or trading
      - Pure data/analytics without control capability
      - Forecasting services without on-site hardware or control
    see_also: Markets & orchestration (when monitored assets participate in markets via VPP)
    subcategories:
      - name: On-site EMS / edge controller
        description: Hardware gateways and edge controllers installed on-site for local energy management — providing near-real-time control of DER (batteries, PV inverters, heat pumps, EV chargers) with offline fallback, protocol translation (Modbus, MQTT, IEC-104), and cloud connectivity. Also covers microgrid controllers and hybrid plant dispatch hardware.
        relevant_for:
          - Edge-deployed hardware for local DER control
          - Protocol translation and device integration (Modbus, MQTT)
          - Offline/fallback operation capability
          - Gateway for cloud-to-site connectivity
          - Microgrid and hybrid plant dispatch hardware
          - Multi-commodity optimisation (electricity + heat + hydrogen)
        belongs_elsewhere:
          - Cloud-only software VPP → Markets & orchestration > Utility & C&I VPP Aggregation
          - SCADA servers for utility-scale plants → SCADA
      - name: SCADA
        description: Industrial-grade supervisory control and data acquisition systems for utility-scale power plants — providing real-time HMI dashboards, alarm management, event logging, remote control, automated reporting, and grid management capabilities including power flow forecasting and congestion management.
        relevant_for:
          - Utility-scale plant supervisory control
          - Real-time HMI dashboards and alarm management
          - Event logging and compliance reporting
          - Remote plant control and configuration
          - Distribution-level DER visibility and grid congestion management
          - Redispatch 2.0 and regulatory compliance
        belongs_elsewhere:
          - Cloud-based portfolio monitoring → Asset performance management
          - Grid-code compliance and power curtailment → Power plant controller (PPC)
          - Building-level energy management → Building & C&I energy management
      - name: Power plant controller (PPC)
        description: Grid-side controllers that regulate active and reactive power output of generation and storage plants to meet grid-code requirements — covering voltage/frequency regulation, ramp-rate control, curtailment management, and grid compliance for utility-scale solar, wind, and BESS plants.
        relevant_for:
          - Active and reactive power regulation
          - Grid-code compliance and certification
          - Voltage and frequency support
          - Ramp-rate control and curtailment
        belongs_elsewhere:
          - Behind-the-meter power management → Building & C&I energy management
          - Market-side power trading → Markets & orchestration
          - Fleet monitoring dashboards → APM
      - name: Asset performance management
        description: Cloud-based fleet monitoring and analytics platforms for multi-site renewable portfolios — providing centralised dashboards, KPI tracking, performance benchmarking, and smart alarms. This category is strictly for pure-play monitoring software and excludes providers that also supply SCADA/PPC hardware, or fully integrated commercial asset management (ERP/CMMS).
        relevant_for:
          - Multi-site portfolio dashboards and KPI tracking
          - Performance benchmarking and loss categorisation
          - Centralised energy data management APIs
        belongs_elsewhere:
          - Dedicated predictive maintenance AI / unsupervised learning → Predictive maintenance & AI analytics
          - Full technical and commercial asset management (ERP/CMMS) → O&M / CMMS
          - Platforms including local SCADA or PPC hardware → SCADA
          - Battery-specific health diagnostics → Analytics > Battery analytics & diagnostics
          - Investment-case modelling → Analytics > Investment & asset valuation
          - Pure weather data, resource assessment, or forecasting → Analytics > Weather & resource assessment
          - Route-to-market trading, BESS optimisation, or VPP dispatch → Markets & orchestration
          - Green hydrogen production optimisation → out of scope
      - name: O&M / CMMS
        description: Comprehensive operating systems for renewable energy operations — combining technical O&M (CMMS, work orders, maintenance scheduling) with commercial asset management (invoicing, contract/warranty tracking, direct marketing, and financial reporting).
        relevant_for:
          - Maintenance scheduling and work order management (CMMS)
          - Contract, warranty, and spare parts inventory tracking
          - Commercial asset management and financial reporting
          - Field-service dispatch and coordination
        belongs_elsewhere:
          - Pure performance analytics and dashboards without workflows → Asset performance management
          - Building facility management → Building & C&I energy management
          - Predictive maintenance algorithms → Analytics > Predictive maintenance & AI analytics
      - name: Building & C&I energy management
        description: Behind-the-meter optimisation platforms for commercial buildings, industrial sites, and multi-site portfolios — covering HVAC control, EV charging management, self-consumption maximisation, peak shaving, dynamic tariff optimisation, and energy hub coordination.
        relevant_for:
          - Commercial building energy optimisation (HVAC, lighting)
          - Self-consumption maximisation and peak shaving
          - Dynamic tariff optimisation and smart charging
          - Multi-site energy hub coordination
          - District heating/cooling grid optimisation
        belongs_elsewhere:
          - Energy efficiency analytics without active control → Analytics > Energy efficiency analytics
          - Flexibility monetisation on markets → Markets & orchestration > Demand Response & Load Flexibility
          - Utility-scale plant monitoring → SCADA or APM