engineering insights
defensible techno-economic models, capital analysis, and bespoke tools for energy, climate, and infrastructure teams
the seat between engineering and finance
I build defensible models, capital analysis, and bespoke tools for energy, climate, infrastructure, and industrial teams. Most useful when the problem is too analytical for a general operator, too commercial for a pure engineer, and too messy for a standard consultant.
what I do
make the economics defensible
Techno-economic models, diligence reads, and capital analysis where the assumptions have to hold up in front of a board, lender, investor, or regulator.
turn messy work into tools
Internal apps, dashboards, data pipelines, and workflow systems that replace manual assembly with something the team can keep using.
make ai useful where mistakes matter
Document intelligence, agents, and automation for regulated or technical work, with sources attached, approval gates where needed, and audit trails that survive review.
selected work
valuing first-of-a-kind cleantech
Process-engineering-grade modeling and live valuation for technologies with no reference plants and no easy comparables.
techno-economic valuation
ranking who should decarbonize first
Operator-level emissions economics across hundreds of energy operators, ranking who should spend first and whether it pays.
emissions economics carbon pricing
making clean-energy projects financeable
A renewable-energy capital platform: marketplace, investor and developer portals, corporate dashboards, regulated workflow, and investor-ready economics.
energy finance
document intelligence engine
Turns a mountain of messy documents into a connected map you can question in plain language, every answer cited and checked.
AI data
deal copilot
Reads messy company data, audits the model without editing it, and drafts source-traced deal materials for a boutique merchant bank.
finance systems
private ai, installed and owned
Agents that read documents, route intake, chase receivables, and report across the tools a business already uses, running on hardware it owns.
AI infrastructure
how engagements work
decision sprint
A short, focused read when a board, investor, lender, regulator, or founder needs a clearer answer. The output is a memo, model, or recommendation that can be challenged.
model or tool build
A working artifact: financial model, diligence system, dashboard, API, internal app, or document workflow. Built to be handed over, not babysat forever.
fractional analytical seat
A few months inside the problem when the company needs engineering, finance, and technical product judgment before it can justify a full-time hire.
For small and mid-sized businesses, the same implementation instinct lives in DecarbDesk: AI-enabled automation for scheduling, document processing, customer communication, data extraction, reporting, and the office work everyone knows should have been automated years ago.
selected thinking
the data is the moat
Most enterprise AI pilots fail before the model runs. The data pipeline decides the outcome.
why tokenization usually makes things worse
Tokenization helps only when the same administrative problem repeats hundreds of times.
the minimum viable firm
The cost of serious cognitive work fell, and the minimum team for professional services changed with it.
get in touch
If a project or decision needs a clearer quantitative spine, or if a workflow has become too important to keep running by hand, email [email protected] or connect on LinkedIn.