Lamariah Smith

Attorney + AI Builder · New York

I practice trust law and build AI.
Both, on purpose.

JD, NY Bar, iOS App Store. Same person. Northwestern Law ’24, Spelman PBK. Most recently Private Wealth at Willkie Farr & Gallagher.

See the workConnect
Lamariah Smith

About

A two-track life.

I grew up in Chicago, went to Spelman to study economics, and graduated Phi Beta Kappa before law school at Northwestern. For most of my twenties I’ve been doing two things in parallel: the technical side of private wealth law — trusts, estates, federal tax — and shipping software, including an AI iOS app I built solo and a fintech I co-founded out of undergrad. I think the most interesting work right now is at the edge where those two things meet, and that’s where I’m putting my time.

Projects

What I’ve built.

Four projects across the AI and operations stack — two shipped, one launching, one operational.

  • Marlo AI

    Live

    An iOS app for career interview practice.

    A voice-based AI coach I shipped solo to the App Store. Generates dynamic interview scenarios, gives instant feedback on answer structure, and tracks improvement over time.

    SwiftOpenAIRevenueCatApp Store

  • 20-Agent Build Pipeline

    In production

    A multi-agent system that ships full web apps end-to-end.

    Six tiers, twenty subagents, forward-flowing — strategy, architecture, eight parallel front-end builders, polish, and ship. Production runtime today. Deploys live demo apps to Vercel with measurable quality gates at every stage.

    ClaudeNext.js 16VercelRemotionPlaywright MCP

  • Arena

    Pre-launch

    Two-agent scope-dispute resolution for freelancers.

    An AI debate system — a dev-side agent and a client-side agent argue scope-creep cases to a signable contract amendment. Twenty dollars a debate. MVP launching soon.

    Next.js 16ConvexClerkStripeVercel Workflow

  • Carenova LLC

    Operational

    Lead-gen automation for HVAC and property managers.

    B2B prospecting playbooks built on Apify, Apollo, and LinkedIn. Operational framework with scraper tools, outreach templates, and CRM tracking — packaged for resale.

    ApifyApolloGoogle SheetsNotion

Services

AI work, for law firms.

Four offers, focused. Two-week audits up front so you know what you’re buying before we touch your data.

  • 011–2 weeks

    AI Workflow Audit + Roadmap

    Where AI actually fits in your firm — and where it doesn't.

    A two-week assessment with a written report and walkthrough. Maps your workflows against the current AI stack, identifies the three highest-leverage opportunities, and flags privilege and conflicts risks.

  • 022–6 weeks

    Custom GenAI Tool Build

    Document review, contract abstraction, deposition prep — shipped.

    End-to-end build of a working tool tailored to one of your workflows. Includes prompt design, evaluation harness, deployment, and a training session for the team using it.

  • 033–8 weeks

    Multi-Agent Systems for Legal Work

    Agentic pipelines for research, drafting, and review.

    Same architecture I use to ship six deployed demo apps a week. Multi-step agent systems that coordinate research, draft generation, and quality review with measurable evaluation gates.

  • 042–4 weeks

    Vendor Evaluation

    Harvey, Co-Counsel, Spellbook, Lexis+ AI — picked the right way.

    Vendor-agnostic comparison against your actual workflows, not their demo. Includes an internal prompt library and rollout plan. Pick the right tool before you sign the contract.

All engagements start with a scoping call. Reach out →

Recognition

On the record.

  • 01Northwestern Pritzker School of Law · JD 2024
  • 02New York State Bar · 2025
  • 03Willkie Farr & Gallagher · Private Wealth, alumna
  • 04Spelman College · B.A. Economics · Phi Beta Kappa · magna cum laude
  • 05UNCF Chicago · Millennial Masterminds, Class of 2023
  • 06U.S. Attorney's Office, N.D. Illinois · Summer Intern
  • 07JPMorgan Chase · Markets Summer Analyst

Writing

Slowly, but on purpose.

One anchor essay live, two more on the way. Operator-grade prose, not corporate copy.

Connect

Let’s talk.

For AI engagements, speaking, or to say hello. I read every message myself.