Programme › Curriculum

Applied Sustainability Reporting with Claude · 6 weeks · 12 sessions · 36 hours

The curriculum, week by week

Teaching weeks teach the standard and end with a timed manual exercise. Lab weeks rebuild exactly the same work with Claude — Thursday led from the front by both mentors, Sunday built by you and presented back. One fictional company, Demo Co., runs through all six weeks: the number you calculate in Week 2 is the number you disclose in Week 4 and evidence in Week 6.

Two tracks, running together

Odd weeks you do the work by hand, so you know what right looks like. Even weeks you rebuild the same thing with an agent you wrote the rules for — and that build track escalates.

Week 2 · Your first skill
A SKILL.md in plain English: the method, the sources it may cite, and the rule that makes it stop rather than guess.
What it is
A work instruction, written for a machine instead of a new analyst. No code.
Week 4 · Scoping an agent
An agent.md that names the role and lists exactly what it may read — and what it may never touch.
What it is
A job description plus a system-access badge. Read-only, so it cannot alter your data.
Week 6 · Many agents, one workflow
Four analysts, one per EcoVadis theme, running at the same time — and the orchestration that runs them.
What you keep
The whole .claude/ library. It is yours, it runs again next quarter, and it works on any harness.

We teach this in Claude Code. The method — write the rules down, scope what the agent may touch, verify what comes back — transfers to any tool you already use.

WEEK 1
Teaching
GHG accountingGHG Accounting I · GHG Accounting IIYou build: Your workspace, and a benchmark inventory built entirely by hand.→
WEEK 2
Lab · assessed
Your GHG inventory, with ClaudeBuilding a GHG Inventory with Claude · Participant LabYou build: Your first skill file — ghg-inventory/SKILL.md — grounded in a published factor source, with a refusal rule that makes it stop rather than guess.→
WEEK 3
Teaching
Preparing sustainability reporting using GRIGRI I · GRI IIYou build: A materiality assessment and a content index, done by hand.→
WEEK 4
Lab · assessed
Your sustainability report using GRI Standards, with ClaudeGRI Reporting with Claude · Participant LabYou build: Your first agent definition — what it may read, what it may never touch — plus a citation skill and a red-team skill.→
WEEK 5
Teaching
EcoVadisEcoVadis I · EcoVadis IIYou build: An evidence inventory and a gap register, mapped by hand.→
WEEK 6
Capstone
EcoVadis pack + the integrated capstoneEcoVadis Submission Prep with Claude, and Making the Workflow Reusable · CapstoneYou build: Four agents, one per EcoVadis theme, running at the same time — and the .claude/ library that produced everything, which is yours to keep.→

How it's assessed

Pass mark is 60% in aggregate plus a submitted capstone. Each lab deliverable carries two marks recorded separately — technical correctness from the sustainability lead, iteration quality and workflow design from the AI lead. A flawed deliverable with a sharp, honest critique of where the model went wrong scores higher than a clean one presented without scepticism.

ComponentWeight
Participation and the six take-home tasks15%
Week 2 — GHG inventory deliverable20%
Week 4 — GRI drafting deliverable20%
Week 6 — integrated capstone35%
Peer red-team contribution (Weeks 4 and 6)10%

Mentors

Aayush Anand
Aayush Anand

Sustainability mentor · teaches all twelve sessions

India's 5th SBTi Certified Professional. ISO 14064 lead verifier. 1,000+ practitioners trained. Energy and GHG work across 50,000+ telecom assets. The final word on whether an output would survive an assurer or a buyer audit. Faculty and reviewer for IICA, IIT (ISM) Dhanbad, MNRE and the IPCC — not a training-hours CV.

LinkedIn ↗

Yashaswi Pathak
Yashaswi Pathak

AI mentor · teaches the lab weeks

Head of AI, building production systems on messy operational data — not demos. Published at ICML and AAAI. In the labs you rebuild the same inventory in Claude; he marks whether you caught the invented factor, not whether the prose looked fluent. On the floor through every Sunday lab for 1:1.

LinkedIn ↗ · Google Scholar ↗

Sessions run Thursday and Sunday, 6:30 pm to 9:30 pm IST.

Ready?

Inaugural cohort starts Thursday 12 November 2026. Places are offered by application, read by both mentors. The fee is $720 (20% off $900).

Apply for a seat
Thu 12 Nov 2026 · Both mentors read every application
$900 $720 20% off
Apply for a seat