OurImpactAI.comContact
Sustainability advisory, AI-enabled

The AI-enabled alternative to building an in-house sustainability function.

Deep sustainability expertise — EP&L methodology, luxury and consumer goods supply chains, corporate sustainability strategy — combined with frontier AI tooling most in-house teams can't access. Analysis at a speed and depth conventional consultants can't match, at a cost well below hiring.

For sustainability leads, CFOs, and COOs at mid-to-large corporates whose sustainability function is under-resourced relative to what's being asked of it.

Book the SprintSee the three problems
01 — Physical climate resilience

Climate exposure starts upstream

Most companies model climate risk where they have visibility — distribution, logistics, owned facilities. They miss raw material sourcing, which is where ecosystem depletion and climate sensitivity actually bite. A company can have a well-mapped logistics risk profile and still be blind to the fact that a critical input comes from a single region facing severe water stress. The blind spot comes from how risk assessment is scoped: it follows the org chart, not the material flow.

Why it matters

  • Cost — supply interruption and input price volatility hit margin directly.
  • Timeline — this is a current operating risk, not a 2050 scenario.
  • Compliance — regulators increasingly ask for upstream exposure, not just direct operations.

The approach

Combine geospatial and satellite data, supply chain mapping, and frontier model synthesis to stitch together fragmented supplier and sourcing information and surface exposure patterns a spreadsheet exercise would miss. The insight — that raw material exposure is the underweighted risk — comes from natural capital and EP&L methodology. The tooling makes it tractable at speed.

Illustrative — method, not a delivered client result

The kind of finding this surfaces: mapping raw material sourcing against climate scenarios points to risk mitigation opportunities in the region of 10–15% of annual procurement costs, without adding headcount.

02 — Scope 3

Your Scope 3 roadmap is probably solving the wrong problem

Two failures compound. First, the data is bad and chasing better data is a dead end — suppliers have little incentive to measure rigorously, and what comes back is incomplete and inconsistent. Second, and more important, the leverage usually isn't in the supply chain at all. Emissions are driven by what is asked of suppliers: overproduction, excess inventory, days of inventory carried, logistics timing, merchandising decisions, weak demand forecasting. Pushing suppliers toward green energy yields something, rarely a lot. Fixing your own planning and production discipline yields more, and reduces working capital at the same time.

Why it matters

  • Regulators want Scope 3 transparency; “our suppliers won't tell us” isn't an answer.
  • Budget spent on supplier engagement programmes with low ceilings is budget not spent on the operational drivers.
  • The operational fixes pay back twice — emissions and working capital.

The approach

Grade and synthesise what supplier data exists, model the gaps with multivariate weighting — production process, company size and maturity, geography, energy mix — then analyse production-to-sales ratios, inventory turns, replenishment cycles, and demand forecasting accuracy to locate the real leverage. Deliver a defensible emissions signature and a reduction plan aimed at the drivers that actually move it.

Illustrative — method, not a delivered client result

The kind of finding this surfaces: a consumer goods business assumes its Scope 3 problem is supplier energy. Production-to-sales ratio and inventory turn analysis traces a majority of the footprint to carrying excess stock and inefficient replenishment. The fix reduces emissions and frees working capital.

03 — Regulatory velocity

Regulation is moving faster than your ability to respond

Climate disclosure requirements land continuously — CSRD, ISSB, California, SEC, and whatever follows — each with different scope, data requirements, and timelines. Three failures stack: teams don't know with confidence which rules apply to them, they can't map requirements to their actual operations and data, and they don't have the reporting infrastructure or the bandwidth to build it in the window available.

Why it matters

  • Regulatory risk with real consequences — missed deadlines and misstatements carry teeth.
  • Late or thin disclosure invites investor and stakeholder questions.
  • Every quarter of delay compounds against the next wave of requirements.

The approach

Synthesise the regulatory landscape against a specific company footprint — sector, geography, structure, size — to determine what genuinely applies. Map requirements against existing data and processes to identify gaps before an audit does. Sequence a phased build rather than a full infrastructure rebuild. Frontier models compress work that would otherwise take a compliance team months.

Illustrative — method, not a delivered client result

The kind of finding this surfaces: a multinational manufacturer facing several new disclosure regimes in the same window, with no reporting infrastructure, reaches a phased compliance roadmap in weeks rather than a six-month scramble — without hiring a dedicated compliance officer.

Red Team

Someone is going to red team your sustainability report. Better it's us.

Adversarial review of sustainability reporting, claims, and reputational exposure.

The problem

Sustainability reports are written by the people who are proud of them, reviewed by the people who commissioned them, and assured against process rather than plausibility.

Then they're read by regulators, NGOs, activist investors, short sellers, journalists, litigators, and ESG raters — none of whom are reading generously.

The gap between those two readings is where reputational and legal risk lives. Most companies never see it until it's public.

Why now

The asymmetry that protected sloppy reporting has collapsed.

Until recently, cross-referencing a 200-page sustainability report against a company's financial filings, procurement disclosures, press coverage, supplier lists, satellite imagery, and prior-year claims required a research team and months of work. Almost nobody did it.

Frontier AI has made that an afternoon's work for a single motivated analyst. Campaign groups have it. Journalists have it. Plaintiff firms have it. Competitors have it.

Your report is being read more adversarially than it has ever been, by readers with more capability than they have ever had. The only question is whether you see what they'll find first.

What we look for

Unsupportable claims

Statements the underlying evidence doesn't carry, particularly forward-looking targets and reduction claims.

Internal contradiction

Where the sustainability report disagrees with the annual report, investor materials, procurement policy, or last year's disclosure.

Methodology exposure

Boundary choices, baseline resets, restatements, and offset treatment that a hostile reader will characterise as convenient.

The Scope 3 problem

Reduction claims resting on supplier engagement programmes with low ceilings, or on estimates that won't survive scrutiny.

Omission risk

What a reasonable reader would expect to find, and doesn't.

Rater mechanics

How the disclosure will actually score against the frameworks and raters that matter in your sector, and where cheap points are being left behind.

Regulatory exposure

Where claims sit relative to green claims rules, disclosure regimes, and current enforcement posture.

How it works

We ingest the report alongside everything else publicly attributable to the company — filings, prior disclosures, investor communications, press, supplier and procurement data, geospatial sources where relevant.

Frontier models do the cross-referencing at a scale and speed no review team matches. Human judgement determines which findings actually bite, who would raise them, and how a company survives them.

The deliverable

A ranked findings register. Each issue, the reader most likely to raise it, the likely framing, severity, and a recommended response — correct, substantiate, reframe, or prepare to defend.

Timeline: two to four weeks depending on scope.

Run it under privilege

Red team findings are written evidence that a company knew about a weakness. In a disputed or litigated matter, that document may be discoverable. We can run the engagement through your legal function or external counsel under privilege. Most clients should. We raise it because you shouldn't have to.

Independence

Red teaming a strategy we designed isn't red teaming. Where we've advised on the underlying roadmap, we disclose it and scope the review accordingly.

Book a scoping callSee what a hostile reader finds

The AI Readiness Sprint

Productised entry point. Fixed scope, fixed fee. Two half-day sessions, two weeks apart — sustainability team and IT in the room together, IT from session one, not brought in at the end to veto.

01

Session one — diagnostic

Where is the team spending time a frontier model could compress? Where is Copilot or equivalent hitting its ceiling? What data can and cannot leave the building? Output: a shortlist of candidate use cases.

Between
Two or three of those use cases are run on frontier tooling — real output, on their problem, without their data.
02

Session two — demonstration and roadmap

What came back, where it beat the current process, where it didn't. Then a prioritised 90-day plan.

The deliverable

A written roadmap: use cases ranked by value and feasibility, data governance constraints mapped, and a build-versus-partner recommendation for each.

From $18K

Progression: Sprint → 90-day pilot on a single use case → quarterly retainer.

Data privacy and governance

Your data stays where it is.

The most common blocker to frontier AI in a sustainability function isn't capability — it's the legal and IT question of what leaves the building. Engagements are structured so that question doesn't have to be answered up front.

We work from problem statements, not datasets

You describe the question. Demonstrating what frontier models can do with a problem of that shape doesn't require your supplier contracts, emissions ledger, or production data.

Demonstrations run on public, synthetic, or aggregate inputs

Industry benchmarks, published disclosures, geospatial and satellite data, anonymised proxies constructed to match your profile. You see the method and judge the quality, then decide whether and where to run it on real data.

If real data comes into scope, governance is mapped first

Which data classes can move and which can't, and what a compliant architecture looks like — on-premise, private deployment, or partner tooling inside your own environment. That mapping is part of the roadmap deliverable.

IT is in the room from the start

Most enterprise AI adoption stalls at exactly this point: the pilot works, legal reviews it, it dies. Governance before deployment isn't caution — it's the only route that reaches production.

Honest boundary

Some analyses genuinely require real data. Deep supplier-level Scope 3 modelling on an actual ledger needs the actual ledger. Where that's the case, the governance track in the roadmap is the path — real work, scoped and priced as such.

Training

We teach this because we do it.

Training isn't a side business. It's how the expertise transfers, and it's where most engagements begin. The demand has been inbound — running EP&L at Kering, work with Moncler, and current depth in frontier tooling is an uncommon combination.

Executive briefings

In-person, regional. For leadership teams who need to understand what's arriving and what it means operationally.

Delivered for executives in Saudi Arabia; a US executive programme requested.

Online courses

Supply chain resilience, delivered through doepicgood, a Dubai/UK sustainability training platform. Scalable, self-paced, international reach.

Live course.

In-house workshops

Custom sessions bringing sustainability and IT teams together around AI adoption, frontier tooling, and what's actually possible versus what the current stack allows.

Delivered for a global pharmaceutical manufacturer.

Speaking

A keynote creates the recognition that there is a problem.

Three signature talks, mapped to the three problems, so a booking is also a positioning statement.

  • Your climate risk model stops too early

    Why exposure assessment follows the org chart instead of the material flow, and what sits in the gap.

    Boards, risk committees, operations leadership

  • Most Scope 3 plans are aimed at the wrong lever

    Supplier energy versus overproduction, inventory, and demand forecasting.

    Sustainability conferences, industry associations, retail and consumer goods

  • What frontier AI actually changes for sustainability teams

    Capability, limits, and the governance question that stalls adoption.

    Executive forums, AI-and-industry events

Past engagement list in preparation — real or labelled, same rule as everywhere else.

Perspectives

Conversations, not interviews.

The hook is always the guest's specific angle — what they're seeing that others aren't, where the friction is, what frontier AI changes in their world. One a month, structured by the three problems.

Two chairs set up for a recorded conversation in a quiet studioIn progress

Neil Brown — the investor-side view

Socially responsible investment fund management; Stanford AI coursework. Climate risk and AI from the capital side.

Syndication

Full conversation on site → three-minute clip on LinkedIn → sixty-second teaser → written pull-quote. One conversation yields four or five pieces. Consistent beats voluminous: one a month sustained outperforms five in a burst.

About

Most AI-and-sustainability advisors know one side or the other. This practice sits at the intersection.

Environmental Profit & Loss methodology run at Kering, engagement with Moncler, and natural capital work across luxury and consumer goods supply chains — combined with current, practical depth in frontier AI tooling. The evidence that the intersection is real is the training demand arriving unsolicited.

  • AI enablement session and workshop facilitation for a global pharmaceutical manufacturer (sustainability + IT)
  • Advisory work on AI-accelerated sustainability ratings and market expansion
  • Executive training delivered in Saudi Arabia; US programme requested
  • Supply chain resilience course via doepicgood
  • EP&L work at Kering; Moncler engagement

Start with the Sprint.

Describe the problem, not the dataset. A short note is enough to establish whether this is the right fit.