AERARIVMThe treasury

Capital & Investment

AI now shapes real-asset demand and the work of investing. We train investment committees, deal teams and operating partners to test AI-driven assumptions and to apply AI from sourcing to exit.

IThe landscapeAI for the deal team, the committee and the portfolio.

The landscape

Real-asset investors underwrite power plants, pipelines, buildings, networks and the companies that run them. AI now affects both the assets and the work of investing in them. It shapes demand theses, from data center power to logistics, and it changes how data rooms are read, models are built and portfolios are monitored. Arclight works with investment committees, deal teams and operating partners in private equity, infrastructure funds and family offices. We teach them to apply AI across sourcing, diligence, underwriting and value creation, and to test the AI assumptions inside their own investment cases.

  1. I

    An AI assumption in every thesis

    Data center demand, power prices, grid capacity and labor productivity now sit inside real-asset investment cases. Investment committees need the fluency to test those assumptions rather than adopt them from a sell-side model or a vendor.

  2. II

    Distributions and value creation

    Exit markets have been uneven, holding periods have lengthened and limited partners are pressing for distributions. They also want evidence of operational value creation, including disciplined use of AI across funds and portfolio companies.

  3. III

    Speed without shortcuts

    Competitive processes compress diligence timelines while data rooms grow larger and more technical. Deal teams must review contracts, permits, technical reports and models faster without lowering the standard of the IC memo.

IIApplications06

Where AI earns its place.

Six applications with proven economics in this sector, taught with the techniques behind them and the ways they fail.

  • 01

    Sourcing and screening

    Scan filings, permits, interconnection queues, ownership records and news to identify assets and companies before they reach a formal process. Screening models rank targets against the fund’s mandate and return thresholds.

    • NLP on filings and news
    • Entity resolution
    • Learning-to-rank models
  • 02

    Data-room diligence

    Extract and compare key terms across PPAs, offtake agreements, leases, concessions, permits and debt documents, with every finding linked to its source. Teams review exceptions and red flags, and carry cited findings into the IC memo.

    • Document intelligence
    • Retrieval-augmented generation
    • NLP on contracts
  • 03

    Underwriting and scenario analysis

    Build probabilistic views of price, volume, availability, cost and exit value, and test them against rate, policy and demand scenarios. The investment committee sees a distribution of returns, not only a base case.

    • Monte Carlo simulation
    • Bayesian inference
    • Time-series forecasting
  • 04

    Physical climate and site risk

    Assess flood, wildfire, heat, subsidence and water stress at the asset level from geospatial and satellite data. Findings feed insurance, capex and valuation assumptions rather than sitting in a separate appendix.

    • Geospatial ML
    • Satellite and SAR imagery
    • Catastrophe modeling
  • 05

    Portfolio monitoring

    Consolidate KPIs, covenants and operating data across portfolio companies and assets, and flag deviations from plan early. Operating partners spend their time on exceptions, and LP reporting draws on one controlled source.

    • Anomaly detection
    • Time-series forecasting
  • 06

    Value creation in portfolio companies

    Identify and deliver AI use cases inside portfolio companies, from predictive maintenance to pricing and procurement, with value tracked against the investment plan. At exit, the evidence is ready for a buyer’s diligence.

    • Predictive maintenance models
    • Price optimization
    • LLM agents
IIIThe sector track04 modules

The Capital curriculum.

Taught within the Command Intensive, as an on-site immersion, or privately in one-to-one advisory.

  1. I

    AI as an investment theme

    How AI changes demand, pricing and risk across power, data centers, networks, real estate and industrial assets. Participants learn to separate durable demand from speculative assumptions and to test both in the model.

  2. II

    AI in the deal process

    Applying AI to sourcing, diligence and underwriting while preserving the quality and auditability of the IC memo. Covers confidentiality in data rooms, model validation and the limits of generated analysis.

  3. III

    AI in the portfolio

    Running AI value-creation programs across portfolio companies: selecting use cases, staffing them, tracking value and reporting it to LPs. Includes how to diligence the AI claims of a target or a vendor.

  4. IV

    Fund governance and disclosure

    Firm-level AI policies, LP questionnaires, regulatory expectations and disclosure discipline, including the risk of overstated AI claims. The module ends with a firm-wide roadmap and clear ownership.

Who attends

  • Managing partners and investment committee members
  • Chief investment officers
  • Deal partners and principals
  • Operating partners and portfolio operations teams
  • CFOs and heads of investor relations
  • Family office principals and investment directors

Outcomes

  • The fluency to test AI-driven demand assumptions in any investment case
  • A faster diligence process that keeps the IC memo to standard
  • A portfolio value-creation playbook with tracked, auditable results
  • Firm-level AI governance ready for LP and regulatory questions

Begin

A sector briefing for your leadership, within seventy-two hours.

Tell us about your portfolio and the decision in front of you. A principal replies within one business day.