Daniel Vetsko's climate portfolio, Kith Climate's Climate Work Residency

Six weeks of climate work for a mid-market private equity fund

Carbon inventory, materiality, physical risk and LP (investor) disclosure for Northwood Capital, built end to end with AI and checked like an audit. Fictional client; real method.

I turn a fund's scattered climate data into numbers a buyer's diligence team can check.

21,778 t CO2eScope 1 and 2 (direct emissions plus purchased energy), calendar 2025, over the months Northwood owned each company (23,437 t for the full year)
$12.6MPhysical-risk expected loss over each company's hold, gross of insurance (range $4.0M to $26.8M); likely understated, since routine weather already in earnings is left out
0 of 10Companies with board climate oversight, a climate role or a target: a gap a buyer will ask about

Each figure's basis is on its week page; physical-risk figures are never added together.

Six weeks, one chain

  1. Week 1, Inquiry and pitch: Read the brief back. Then answer it. 13-slide proposal; $220,000 fixed fee (illustrative). Passes to Week 2: Promised: a data audit.
  2. Week 2, Data audit: Ten companies. None ready for a buyer's diligence. Readiness 26 to 38 out of 100. Passes to Week 3: 10 companies, 138 facility rows.
  3. Week 3, Carbon inventory: A carbon number a buyer can check. 21,778 t CO2e over Northwood's ownership period (23,437 t for the full year). Passes to Week 4: Each company's footprint.
  4. Week 4, Materiality: 46 exposures. 5 worth acting on. About $117K answers customers' carbon-data requests. Passes to Week 5: Exit timing and hold years.
  5. Week 5, Physical risk: What weather could cost before each exit. $6.04M a year expected loss, likely understated. Passes to Week 6: Loss over the hold, and its method.
  6. Week 6, Disclosure: A disclosure Northwood can rerun without us. LP report, ten board pages, re-run steps.
  7. AI builds fast. Separate checks make it right. Rules written before results; Read-only reviewer agents; Checks that must prove they can fail; Independent recalculation; Gates that freeze what was signed.

Six weeks, one client. Each week built on the last.

Week 1, Inquiry and pitch

Read the brief back. Then answer it.

13-slide proposal; $220,000 fixed fee (illustrative)

Passes to Week 2 Promised: a data audit

Week 2, Data audit

Ten companies. None ready for a buyer's diligence.

Readiness 26 to 38 out of 100

Passes to Week 3 10 companies, 138 facility rows

Week 3, Carbon inventory

A carbon number a buyer can check.

21,778 t CO2e over Northwood's ownership period (23,437 t for the full year)

Passes to Week 4 Each company's footprint

Week 4, Materiality

46 exposures. 5 worth acting on.

About $117K answers customers' carbon-data requests

Passes to Week 5 Exit timing and hold years

Week 5, Physical risk

What weather could cost before each exit.

$6.04M a year expected loss, likely understated

Passes to Week 6 Loss over the hold, and its method

Week 6, Disclosure

A disclosure Northwood can rerun without us.

LP report, ten board pages, re-run steps

How it was checked

AI builds fast. Separate checks make it right.

  1. Gates that freeze what was signed
  2. Independent recalculation
  3. Checks that must prove they can fail
  4. Read-only reviewer agents
  5. Rules written before results

Daniel Vetsko

I turn a fund's scattered climate data into numbers a buyer's diligence team can check.

Pick any week on the line to see the work.

Climate Work Residency, Kith Climate. Fictional client; real method.

The work, week by week

Week 1

Inquiry and pitch

Northwood asked what a six-week, AI-native climate engagement would deliver, and which of its ten companies should come first.

See the work
Week 2

Data audit

Is Northwood's climate data good enough to disclose to LPs (the fund's investors) and to survive a buyer's diligence, and what must be fixed first?

See the work
Week 3

Carbon inventory and ESG scorecard

What are the ten companies' Scope 1 and 2 emissions (direct emissions plus purchased energy), can Northwood defend the number to a buyer or LP (fund investor), and what should it act on?

See the work
Week 4

Materiality and initiatives

Which climate exposures at Northwood's ten companies actually move what a buyer pays, and what should Northwood do about them first?

See the work
Week 5

Physical climate risk

What could hurricanes, floods, drought, heat and storms cost Northwood's ten companies before each exit, and what should Northwood do about it?

See the work
Week 6

Disclosure as a service

How can Northwood answer climate questions from LPs (its investors) and buyers now, and keep doing it next year without Kith?

See the work

Skills shown

How I built it

Alongside the client work I built a way of working that makes AI-produced analysis checkable: standing instructions for the AI, read-only reviewer agents, rules written before the results, checks that must prove they can fail, and gates that freeze what was signed. The capstone tells how that grew from no review in Week 1 to gated, independently checked delivery by Week 6.

Read how I built it with AI

About this portfolio

This is my work from the Climate Work Residency, a six-week programme run by Kith Climate. Northwood Capital and its ten portfolio companies are a fictional case, and the data room is Kith Climate's; the analysis, the deliverables and the way they were built are mine, made with Claude Code.