AppliedESG.ai

Syllabus

Ten modules in four parts: the foundations of working with AI, the core craft of directing it, the guardrails that make its results audit-ready, and AI’s applications and environmental impacts in ESG work.

Modules 1–3

Foundations

  1. How AI works, from generative to agentic

    How generative AI produces answers, and what changes when an agent can use tools, work with your files, and complete multi-step tasks on its own. Choose which tasks to start with.

  2. How AI fails (and what to do about it)

    How AI errors differ from a junior analyst’s: fabricated sources, silent gaps, calculation and unit errors, misread documents, and meaning lost in summary. The controls that prevent them, the quick checks that catch them, and how to fix the cause so they don’t recur.

  3. What the AI sees, and what it costs

    What it knows about your work, and what it’s quietly carrying over from earlier sessions. How to deliberately supply source material, and how to choose input and output formats that are more reliable and more token-efficient.

Modules 4–6

The core craft

  1. From chatbot to agent: briefing AI like a colleague

    Specifying the task, the audience, the methodology, and the constraints, and saving those instructions so you don’t start from scratch each time.

  2. Judgment vs. calculation: when to make AI show its math

    When to rely on the model’s inference to make a judgment call, and when to force it to act like a computer by writing and running a script so that the result is reproducible and auditable. Matching the tool and the depth of reasoning to the task.

  3. Designing reusable multi-step workflows

    Breaking a large job into stages, using MCP connectors, and adding human approval points where needed. Coordinating a lead agent with sub-agents that each handle part of the work. Packaging a workflow as skills and connectors so it can be run by a colleague or on a recurring schedule.

Modules 7–8

Building the guardrails

  1. Verification and traceability for audit readiness

    Designing checks proportionate to risk, and leaving a “paper trail.” Run-time checks: recomputation, reconciliation, repeatability, critic agents, and known-answer tests. Traceability: audit trails, figure-level lineage, and readiness for assurance or internal audit. Keeping it reliable over time: defect logs, metrics, and re-validation when tools change.

  2. Governance, security, and boundaries

    What data can go into which tool, and who decides. Working with internal AI stakeholders, such as IT, legal, and procurement. Recognizing prompt injection, and limiting agent access to minimize exposure. Deciding what should never be delegated.

Modules 9–10

AI for ESG

  1. The environmental cost of AI compute

    What happens in the data center each time you send a prompt, and how to account for it. Energy, water, and emissions across model training and usage, and where they belong in a GHG inventory. Why published per-token or per-query estimates vary so widely. Practical levers for shrinking the footprint and weighing it against the work AI replaces.

  2. ESG workflows in practice

    Creating responses to EcoVadis, CDP, S&P Global CSA; building an RFP response workflow; creating a regulatory tracker; reviewing changes to standards, such as GHG Protocol & SBTi; AI-assisted sustainability reporting; and more.

Outcomes

What you'll be able to do

  • Draft questionnaire responses

    Answer CDP, EcoVadis, and customer surveys from approved text, and find quick score gains.

  • Defend the work

    Document the workflow and provenance for your internal audit team or an assurance provider.

  • Map narratives and KPIs to compliance language

    Connect how your company describes its work to what each standard, rating, or regulation requires, and record the reasoning.

  • Translate ESG for other company teams

    Brief sales on customer requirements, legal on claims, finance on disclosure controls, and IT on data governance.

  • Support your GHG inventory

    Structure utility, travel, and spend data, catch unit errors and outliers, and reconcile against prior years, with arithmetic you can inspect.

  • Run a standing regulatory tracker

    An agent that watches CSRD, ESRS, ISSB, and sector guidance, and explains what changed.

  • Benchmark peers in a day

    Pull a dozen peer reports into one comparison table with page-level citations.

Interested in joining the workshop?

The first workshop dates are coming soon. Sign up to hear first and get the early-bird rate.

Sign up