Case study · Fintech + Healthtech · Solopreneur economy
Engagement: Embedded technical partner · growth-stage scaling, ongoing
Lettuce is a tax automation, bookkeeping, S Corp, and embedded-banking platform built for solopreneurs — content creators, real estate professionals, therapists, consultants, healthcare practitioners, and other self-employed workers earning $100K+. Since Stackpad embedded with the team, the platform has grown from a single tax-automation product into a multi-product company: automated bookkeeping and continuous tax optimization, S Corp formation and management, business banking, an AI assistant for tax questions, and — most recently — healthcare. Stackpad has been the embedded engineering partner across that expansion, including the move from fintech into the regulated healthtech surface that comes with it.
Result
Scaled tax-filing and bookkeeping workflows through sustained customer growth
Result
Supported the expansion from a single product into banking, an AI assistant, and healthcare — without rebuilding the foundations underneath
Result
Engagement remains active — still embedded as Lettuce continues to grow
The constraints
Multiple money-adjacent workflows — tax filing, bookkeeping, and bank-partner integrations — had to scale concurrently, with no tolerance for off-by-one errors in financial calculations.
MVP-era code carried assumptions that quietly broke as customer volume grew, especially in the financial calculation paths that auditors, banks, and customers all eventually read.
Tax-season load is unforgiving — peak volume turns minor latency or correctness issues into immediate customer crises.
The platform kept adding product surfaces — S Corp formation, business banking, an AI assistant, healthcare — while the existing ones were still scaling.
Moving into healthcare added a second regulated domain on top of the fintech one: new partners, new compliance expectations, and sensitive data under different rules — all while the financial side kept growing.
Bank-partner relationships (BaaS, debit-card issuance, ACH) carry their own SLA expectations, audit posture, and operational requirements distinct from core product engineering.
What we shipped
01
Hardened the systems that absorb tax-season volume spikes — the financial calculation paths, the queueing and retry behavior, and the integration code with downstream partners. Reduced the operational toil that comes from fragile MVP-era assumptions surviving past their useful life.
02
Re-worked the parts of the code where money math, accounting integrations, and tax logic all collide. Made the calculation paths legible to engineers, auditors, and customer-success teams who all read them for different reasons.
03
Cleaned up the partner-integration surface as the platform expanded from one product into many — banking, S Corp services, healthcare. Made the integration code behave consistently regardless of which downstream partner was on the other side.
04
Established the operational controls that high-volume, money-adjacent platforms need — observability into the financial flows, replay tooling for failure recovery, and the kind of evidence trails fintech operators end up needing whether or not they’re in an active audit.
Outcomes
Currently underway
Stackpad remains embedded with Lettuce as the platform continues to expand. Lettuce has added healthcare to a product surface that already spanned tax, bookkeeping, S Corp services, banking, and an AI assistant — pushing the platform across the line from fintech into regulated healthtech, where money-adjacent and health-adjacent data live under different rules at the same time. The work continues along the same fault lines — reliability under growth, financial-calculation correctness, partner-integration posture, and the operational controls a multi-product, multi-regulated platform needs to keep adding surfaces without rebuilding its foundations.
If you’re running a fintech that’s outgrowing its MVP assumptions — or expanding into new product surfaces while the existing ones still need scaling — let’s talk.
