Jarvis Labs — by Adrian Kay

Hi, I'm Adrian.
I build the AI that runs real-estate operations.

Slow follow-up loses deals. Manual underwriting eats the week. I build AI for real estate — deal flow, underwriting, and CRE data — the same institutional-grade system I run, installed in yours.

Book a callSee what you get

Real estate · Engineering · Professional services

What you get

One system for the work that’s costing you deals.

  • 01

    ML & Data AnalyticsSignature

    Investment-ranking models, forecasting, and the pipelines that feed them — institutional-grade.

  • 02

    Process Optimization

    Find where the hours leak — then close the leaks for good.

  • 03

    Systems That Run Themselves

    The repetitive work handled end-to-end, so your team stops touching it.

  • 04

    CRM Setup & Integration

    HubSpot wired to your deal flow. Clean data, no double entry.

  • 05

    Knowledge Systems

    Searchable memory across documents, deals, and comms.

  • 06

    Speed-to-Lead Agents

    Instant, qualified replies so no inbound goes cold.

  • 07

    Custom AI Skills

    Agents tuned to your underwriting and screening — not generic chatbots.

Fluent in the institutional data stack

AxiometricsiLEVEL

About

Engineering-grade AI, proven inside institutional real estate.

I trained as a civil engineer at Arup, designing and building real projects in Hong Kong — and automating the tedious parts with Python and VBA before that was normal.

For the last six years I’ve been a senior data scientist at Madison International Realty, building the ML deal-ranking models, ETL pipelines, and analytics platform a global real-estate PE firm actually runs on. This is the day job, not a theory.

What’s rare is the combination — civil-engineering discipline, hands-on design-and-build experience, and production data science, right down to the GPUs and networking in my own self-hosted AI lab. Jarvis Labs brings that in-house capability to operators who can’t staff a data-science team.

Adrian Kay

Adrian Kay

Founder, Jarvis Labs

Stack

AI & ML
ClaudeClaude CodeCoworkCodexMLflow
Data & Analytics
PythonSQLAirflowAWS RedshiftMicrosoft FabricPower BIPostgres
Build & Ops
n8nNext.jsHubSpot

Track record

  1. Founder

    Jarvis Labs
    2026 — Present

  2. Co-founder

    Vantium AI
    2026 — Present

  3. Senior Data Scientist

    Madison International Realty
    2020 — 2026

  4. Data Scientist

    Megalith Capital Management
    2019 — 2020

  5. Civil Engineer

    Arup
    2015 — 2019

Education

  • MEng, Civil Engineering · University College London (UCL)
  • Data Science Fellow · NYC Data Science Academy

Engagement

Three ways to work together.

  1. Operations Audit

    Fixed fee · credited toward a build

    A paid, fixed-scope read of your operation.

    • Impact matrix + roadmap
    • Tools matched to your operation
    • Yours to keep, build or not
    Start with an audit

  2. Build SprintStart here

    Scoped per project

    I implement the roadmap in focused sprints.

    • Deal-flow, CRM + inbox systems built
    • Speed-to-lead & underwriting agents
    • Built on tools you own
    Scope a build

  3. Operate

    Monthly retainer

    I run and extend the stack as you grow.

    • Monitoring + upkeep
    • Continuous extension
    • Priority support
    Talk retainer

Every engagement starts with the audit.

Questions

Before you book.

Get started

Let’s find the hours you’re losing.

A 30-minute working session. You leave with a concrete next step — whether or not we build together.

Book a working sessionak@jarvislabs.dev