Bridge Loans And Direct Equity.A Site As Serious As The Money.
Nine dynamic market pages and AI-powered admin tools for a New York commercial real estate firm doing institutional deals: bridge loans and direct equity across multifamily and medical assets.
The Actual Interface.
These Borrowers Have Seen Every Templated Lender Site.
Commercial real estate borrowers at this deal size are sophisticated. H Equities needed a digital presence that communicated institutional credibility: fast, precise, and without visual noise.
The other cost was origination. Commercial loan outreach is a precision game, and a generic email does not get replied to, so the first draft was eating the team's time on every deal.
Nine Market Pages And A Drafting Tool Behind Them.
SEO architecture. Each asset class, multifamily, medical office, mixed-use, retail and industrial, and each loan product type gets a dedicated, independently optimised page. Deal parameters, LTV ranges, term structures and historical case examples are tailored per page rather than one template with swapped keywords, built to rank on the exact search strings commercial borrowers use.
AI-powered outreach. An internal interface where the team describes a deal, asset type, location, loan amount, LTV and timeline, and the model drafts the initial outreach email to the borrower. Tone-matched to H Equities' voice, relevant deal terms pre-filled, ready to send in under 60 seconds. It replaces the most time-consuming part of the origination workflow.
Structured loan inquiry form. A multi-step intake form captures deal size, asset type, location, timeline and borrower contact, and submits to Supabase. The model automatically generates a response draft for the team, scoped to the specific deal parameters submitted, ready for review and send. No generic auto-reply, no data lost in email.
TypeScript throughout. No runtime surprises on form validation, deal parameter handling or model response parsing. Zod schemas validate every form submission before it touches the database, and the model integration is typed end to end: request shape, response parsing and error fallback are all explicit.
A Draft In Two Minutes, In The Firm's Own Voice.
The drafting tool gives the team a structured way to brief the model and get a first draft that sounds like them.
- Team inputs: asset type, location, deal size, LTV target, timeline
- The model receives a system prompt encoding H Equities' voice and deal criteria
- Output: a first-draft outreach email with subject line, greeting and CTA
- Team reviews, edits inline and sends, total time under two minutes
- Sent emails are logged to Supabase with deal context for future reference
Institutional, not sterile. The visual language targets real estate investors and operators handling eight-figure transactions, who trust precision rather than decoration. Tailwind v4 with a tight token system and no bloat, a near-black, white and single-accent palette, data-table typographic hierarchy instead of marketing copy, and one CTA per page with no scatter.
- Next.js
- TypeScript
- Tailwind v4
- Language model
- Supabase
- Zod
Let’s Build What’s Next.
Bring the business problem. We’ll talk through what would make a difference and where to start.