Cover Image for AI for Real Estate Lending — KV Capital Hackathon
Cover Image for AI for Real Estate Lending — KV Capital Hackathon
Avatar for Open Hackathons
Presented by
Open Hackathons
Hackathons that are open for everyone
Hosted By
47 Going

AI for Real Estate Lending — KV Capital Hackathon

Virtual
Registration
Welcome! To join the event, please register below.
About Event

Build an AI Agent for Real Estate Finance

Solo. 2 weeks to build. $1,000 CAD prize pool. Top finalists fast-tracked to interview for KV Capital's AI Engineer role.


What you're building

KV Capital is a Canadian alternative investment manager with $900M in assets under management. Founded in 2006, we operate across real estate debt, real estate equity, and private equity in operating businesses — the only firm in Canada offering all three under one roof. Within our real estate debt business, every deal we underwrite starts with the same question: what is this property actually worth?

The traditional answer is "comps" — finding comparable recent sales and reasoning through how they relate to the subject property. It's expert work, slow, and full of judgment calls. We think AI agents can do most of it.

Your challenge: build an AI agent that, given a subject property, retrieves and ranks comparable properties to support a valuation estimate — and explains its reasoning. This is one of the highest-leverage problems in our real estate debt business, and an honest preview of the kind of work the AI Engineer role will involve.

The full brief, dataset, and submission instructions arrive via Notion after you sign up.

Prize pool

  • 1st place: $700 CAD

  • 2nd place: $300 CAD

Plus: top 5 submissions by rubric get fast-tracked into the interview pipeline for our AI Engineer role at KV Capital. Win or not — if you're job-curious, this is the shortcut to the front of the line.

How it works

  1. Sign up through this page.

  2. Get the brief. Within 24 hours, we'll send you a Loom walkthrough with the challenge, the dataset, and exact submission instructions.

  3. Build for two weeks. You have 14 days from your signup date to submit.

  4. Submit your GitHub repo and Loom walkthrough.

  5. Finalists announced in the first week of June. Top 5 are invited into the KV Capital interview pipeline.

Who this is for

  • AI engineers and new or recent grads who can ship full-stack

  • Builders comfortable with Python and at least one modern web stack

  • Anyone who's built with LLMs — prompt design, tool use, retrieval, agents

  • People curious about real estate, finance, or fintech (no domain experience required)

This isn't the right fit if you want a team event, a research-heavy ML problem, or something you can knock out in an evening.

Timeline

Once you sign up, you have 14 days to submit. Final cutoff for all submissions is end of May. Finalists announced the first week of June, with interviews starting mid-June.

What you submit

  • GitHub repo (public, or private with access shared) — README with setup, approach, and tradeoffs

  • Loom video (less than 10 minutes) — architecture walkthrough, key decisions, live demo

Both are required. We weight the Loom heavily — clear communication is part of the role.

How we judge

Five dimensions, scored 1–5 by a panel from KV's lending and engineering teams:

  • Code quality and structure — would another engineer want to work in this codebase?

  • Agent and LLM design — tool use, prompt design, reasoning transparency

  • ML and data thinking — comp scoring logic, feature choices, edge case handling

  • Communication — README, Loom, documentation

  • Pragmatism — scope choices and what you cut

Submissions are reviewed in waves as they come in. Blind first pass to filter top 15–20, then full panel review of finalists. Pedigree doesn't enter the rubric. What you've built does.

Effort guidance

Plan for 8–12 hours of focused work within your 2-week window. If you're going much beyond that, scope down — pragmatism is part of the rubric.

The fine print

  • Solo only. No teams.

  • Eligibility. Open globally to enter and win the prize. The interview opportunity is contingent on being legally able to work in Canada.

  • IP. KV Capital as the host of the event owns the full IP.

  • Originality. Use any tools, models, and frameworks you want — including AI coding assistants. Submissions must be original work created for this hackathon.

  • Deadlines are firm. Your individual deadline is 14 days from signup. The overall cutoff is May 31st. No extensions.

About KV Capital

KV Capital is a Canadian alternative investment manager with $900M in assets under management, headquartered in Edmonton with offices in Calgary and Toronto. Since 2006, we have funded over $1.8 billion in investments across three connected lines of business: real estate debt, real estate equity, and private equity in operating companies.

Our tagline is "Capital to compete." We're investing in AI to make our underwriting, portfolio management, and investor reporting faster and sharper than anyone else in our space — and we're hiring an AI Engineer to help build it. This hackathon is one of two streams we're using to find that person.

FAQ

Will the data be real? No — we'll provide a synthetic or anonymized dataset that mirrors the structure of real records.

Can I use Cursor Claude Copilot? Yes. Use whatever helps you ship.

What if I don't want a job? Compete anyway. The prize is yours regardless of whether you interview.

Do I need real estate experience? No. Curiosity matters more than domain knowledge.

When does my 2-week clock start? The moment you sign up. We'll send the Loom brief within 24 hours, but your 14 days starts at signup.

Avatar for Open Hackathons
Presented by
Open Hackathons
Hackathons that are open for everyone
Hosted By
47 Going