SELECTED WORK
Problems we have solved
A selection of engagements across fintech, venture capital, and operating model design. Some ended with a plan to execute. Some ended with a recommendation to stop. Both are the job.
FOUNDER WORK
Daylight
Co-founder & CEO · 2020–2023 · New York
Few people get the chance to build a financial services company, a community institution, and a fix for a systemic problem simultaneously — in a foreign market, mid-pandemic. Daylight was all of those things at once.
The challenges didn't stack neatly. A debit brand competing for attention in a crowded marketplace. A press engine that had to generate cultural goodwill and investor conviction at the same time. A community that had waited a long time for a company like this and held it to standards no mainstream fintech faced. Daylight's lessons were less about the numbers and more about what it takes to thread the needle between problems that are each genuinely hard on their own. Let alone one company at the intersection of all of them.
100k
Sign-ups in approximately 100 days
50%
CAC reduction through integrated multi-channel campaign
30+
Investors including Anthemis, Precursor, Kapor Capital, Gaingels, Restive
$25m
Funds raised in equity and debt financing.
THE THESIS
A world-first problem in one of the world's hardest markets
Everyone has a debit account and standing out in a commodised market is tough. We knew we'd have to go big. We redefined the role a financial institution could play in the life of a community that has never had one designed for them. That is an order of magnitude harder.
We entered the US from Europe with no operating history, conducted discovery interviews across the community, and used what we found to build a founding thesis grounded in evidence. Then we convinced Visa to back an unproven team on an unproven proposition. That commercial relationship made the product possible.
The tasks wasn't feature development, it was market entry, category creation, community research, commercial partnership and regulatory navigation, all from a standing start.
BUILDING THE CATEGORY
Building the team and building the proof simultaneously
The first operational challenge was this: we had to prove demand and raise capital at the same time, using a waitlist as both marketing proof-of-concept and investor data. The product didn't exist yet. The category didn't exist yet. The investors needed to believe in both before they'd write a cheque.
The waitlist was the proof of concept. The campaign was the argument: trucks through queer neighbourhoods, earned media, bars and restaurants, social and influencer. The message — We're here. Stop apologising for being gay — was specific enough to mean something and broad enough to travel. 100,000 sign-ups in 100 days.
A Series A six months after the seed round. We were also managing a continuous low-grade pressure that is hard to describe from the outside: a significant portion of our operating environment — investors, press, some potential partners — could not quite believe a company like ours should exist. We operated under that scepticism for three years and delivered anyway.
THE PIVOT
When external conditions forced a change of direction, the hardest work was not strategic. It was human.
Interest rates moved. We knew it was existential. Keep going and hope, or cut deep: closing a product our team loved, letting go of half the staff, and repositioning into health tech adjacent fertility.
We chose to cut. Deep and fast.
Then came the harder question: how do you bring a team along when they were hired to build an inclusive banking product and you're now asking them to build something entirely different, with half their colleagues gone?
We gave them one thing to believe in: the world's first digital surrogacy and adoption platform for the LGBTQ+ community, subscription-supported and integrated into the bank account. Another world-first - specific enough to mean something. We had limited time and a team that needed a reason to stay. We found one.
INVESTOR MANAGEMENT AND CRISIS COMMS
Building a company under intense media scrutiny
Controversy was the strategy, not the side effect. We courted that attention deliberately to build momentum in an unproven category. It cut both ways.
When it was good, it was very good. When it was bad — every mistake broadcast to an audience that expected better — it required personal resilience and investor trust that can't be manufactured in a board meeting.
The answer was governance. Board members who knew the full operating reality made better decisions than board members who received curated updates. When things went wrong publicly, we could say so clearly — because we'd never had to pretend otherwise internally.
But we weren't accountable only to customers and investors. We were accountable to a community that had never seen a company like ours before and held us to standards no mainstream fintech faced. We were more than a company. We were an institution. And we were judged that way.
OPERATING MODEL DESIGN
Capital Selection & Deployment
European Venture Capital Fund
A European venture capital fund was exploring whether its investment process could be redesigned into something closer to a capital deployment machine : two deals per week, across a diversified index of early-stage startups. The question was whether that ambition was operationally achievable, and what it would take to build it.
HOW WE WORKED
Mapping the operating architecture
The first step was to map how the fund actually operated. Most venture funds understand individual activities — sourcing, diligence, portfolio support — but rarely as an interdependent system. Without that map it is impossible to identify where decision velocity slows, what data is missing, or which parts of the process require judgment versus infrastructure.
We built a complete operating architecture across six layers and defined each as a set of capabilities with structured inputs, outputs, data requirements, and KPIs. We then identified the dependencies that would need to hold for two deals per week to be achievable.
WHAT WE FOUND
The bottleneck was not deal flow. It was decision throughput.
Deals were stalling between initial evaluation and investment committee because diligence was bespoke, narrative-heavy, and impossible to compare across companies. Two structural problems compounded this: deal data arrived as pitch decks rather than structured inputs, and there was no systematic way to detect discrepancies between founder claims at intake and what diligence later verified.
The fund could reach its target velocity, but only if three dependencies were met: a structured intake schema that captured founder claims as a testable baseline, a thesis taxonomy that kept the pipeline genuinely thesis-aligned, and a standardised decision process that separated automated data synthesis from partner judgment.
THE RECOMMENDATION
The model was viable. The fund was shown exactly what building it would require — in data infrastructure, process design, and operating capability. The recommendation was to proceed and raise the capital required to build it.
MARKET ENTRY ASSESSMENT
Healthcare Lending
MEXICO
A fintech founder approached Crucible Hill Partners to assess the viability of a lending product designed to help underserved communities access healthcare procedures. The team had begun building. Before committing further capital, the question was whether the business model could support the ambition.
HOW WE WORKED
Mapping the load-bearing assumptions
We mapped the business plan and identified the two assumptions on which viability depended — the conditions that, if not met, made the strategy unworkable regardless of execution quality:
-
That repayment rates at the planned loan structure would be sufficient to build a healthy book of debt
-
That the product as designed would generate meaningful demand from the target customer segment
We stress-tested both against the financial model. Customer research was not required to test assumption two — stated demand for accessible credit is near-universal and tells you nothing about the economics of repayment.
WHAT WE FOUND
At prevailing Mexican interest rates, the repayment structure was not serviceable for the average target customer. The loan book economics were structurally unsound. The first dependency could not be met. No iteration of the product design resolved it.
THE RECOMMENDATION
Proceed only if the repayment model can be restructured to work at market interest rates. It could not. The recommendation was to stop.
BUSINESS CASE VALIDATION
Neobank Launch
MEXICO
An investor approached Crucible Hill Partners with a plan to launch a neobank in Mexico built around an existing loyalty and community platform. The logic appeared sound: an established user base, prior experience in payments processing, and a plan to fund the venture within the existing P&L.
HOW WE WORKED
Mapping the load-bearing assumptions
We mapped the three assumptions on which the entire business case depended — the conditions that would need to hold for the strategy to be viable:
-
That the existing community platform would deliver low or near-zero cost of customer acquisition
-
That prior payments experience meant compliance infrastructure could be resourced with lower-cost staffing rather than specialist tooling
-
That the venture could be funded within the existing P&L without external capital
For each, we asked: what must be true for this to hold? We then tested each against available data and sector benchmarks.
WHAT WE FOUND
All three dependencies failed.
Day-one active users on the community platform were below 2%. The assumed free acquisition channel did not exist — full market-rate acquisition costs would need to be funded from launch.
Compliance requirements materially exceeded what lower-cost generalist resource could cover. With accurate acquisition costs and real compliance infrastructure, the capital required to reach market was substantially beyond what the existing P&L could absorb.
THE RECOMMENDATION
Proceed only if the business is recapitalised and external investment raised — a fundamentally different proposition from the one presented. The investor's appetite for that path was limited. The recommendation was to stop.