Ship the thing you don't have time to build.
I'm Jake — I design and build AI-backed tools, full-stack web apps, data pipelines, and civic/GovTech systems, end to end and shipped to production. Everything below is drawn live from things I've actually built and written.
Agentic Workflow Design
I install a production agentic workflow in your stack — a multi-step LLM agent loop with a human gate, cost instrumentation, and a runbook. The same machine that runs this site, scoped to one workflow that earns its keep.
What I work in
AI
AI & LLM Engineering
ExpertAgentic pipelines, RAG, prompt design, and LLM-backed tooling shipped in production.
Engineering
Data Migration & Integrations
ProficientREST / SFTP integrations, ETL, and data mapping & validation across enterprise systems (EAM, GIS, CRM) — production-reliable cutovers.
Applied Math & Modeling
ExpertGame theory, optimization, and quantitative modeling turned into working systems.
Full-Stack Web Apps
ProficientNext.js / React apps end to end — App Router, server components, auth, deploys.
Python & Automation
ProficientScripting, data tooling, and automation pipelines in Python.
Data & Backends
Working knowledgeSupabase / Postgres schema design, row-level security, and service-role data layers.
Design
Interactive & Game Design
ExpertEducational and interactive games — Canvas2D / WebGL, mobile-first.
Communication
Research & Technical Writing
ExpertLong-form research, technical essays, and clear documentation.
Training & Enablement
Expert120+ sessions to 650+ users at 4.8/5 satisfaction; role-based tracks and playbooks adopted across 87% of projects.
Delivery
Enterprise SaaS Implementation
ExpertEnd-to-end delivery for enterprise & public-sector clients — scoping, configuration, data migration, UAT, and go-live. 100+ engagements at 92% on-time / 95% on-budget.
Solutions & Sales Engineering
ProficientTranslating business requirements into architecture: technical discovery, deep-dives, and enablement for both technical and non-technical stakeholders.
Customer Success
Technical Customer Success
ExpertStrategic technical advisor — governance frameworks, platform-health monitoring, executive escalation management, and voice-of-customer back to product.
Selected work
Enterprise SaaS implementation track record
5+ years leading end-to-end implementations for enterprise & public-sector clients — discovery through go-live, with a documentation-first, playbook-driven methodology.
100+ engagements · 92% on-time, 95% on-budget · 13 of 14 executive escalations resolved without concessions · playbooks adopted across 87% of projects (rework −28%).
Strata: ten excavations of the work a finished thing hides
An eight-team agent organization that reads this repository as a dig site — the rework, false starts, decisions, and the arcs it keeps circling that a finished website is built to hide. Ten interactive essays and six browsable surfaces, every number excavated from the real commit history.
AI-native product engineering
Designs and ships LLM-powered web apps end to end — Next.js / React / Supabase + Claude, from data model to UI. Example: an AI job-search engine that tailors resumes & cover letters, scores fit, vision-parses PDFs, imports postings, and tracks a hiring funnel.
Full-stack delivery — structured-output LLM pipelines, vision PDF ingest, guarded fetch/import, and a stats dashboard — built and deployed on jakelawrence.xyz.
The Shape of Original
An interactive essay on AI vs human narrative construction, originality as statistical rarity, and the classification infrastructure being poured under authorship.
Seeing Like an AI Company
Six dead scholars read OpenAI's industrial policy proposal and find six connected problems it refuses to name.
Accountability Tracker
25 universities. 11 dimensions. 1.3 million students in scope. The empirical companion to the essays.
The Long Hold
Sixty years of US space nuclear capability, held across six political eras without ever being exercised, until the April 2026 NSTM-3 memo arrived to cash it in.
The Warehouse
The American federal state as a portfolio of 49 unexercised options, evaluated through an acquisition optimizer's lens. What gets cut, kept, monetized, and why.
Stance Design
Every AI system you interact with has been told how to treat you. That decision was made by a product team, encoded in training, and deployed without your knowledge. The essay coins the replacement term.
Fixed-scope, fixed-price
Known scope, known price, paid up front — no proposal dance. Bigger or fuzzier than these? Use the form below instead.
30-minute working session
$150one callA screen-share half hour on your actual problem: architecture read, AI-feature feasibility, debugging a decision — whatever moves you fastest. Booked within a business day; you leave with concrete next steps.
Hosted screenshot uploader
$19one-timeA personal upload token for the jakelawrence.xyz ShareX endpoint: EXIF-stripped hosting on a real CDN with branded short links, plus a ready-to-import .sxcu config. Delivered to your email within a minute of payment.
Supabase schema & RLS review
$500fixed scopeI read your schema, policies, and access paths end to end and deliver a written report: concrete RLS gaps, indexing and shape problems, and the migration order to fix them. Turnaround inside a week.
Next.js / AI app audit
$750fixed scopeA production-readiness audit of your Next.js or LLM-backed app: architecture, server/client boundaries, prompt and tool-call design, cost hot spots — delivered as a prioritized findings doc plus a call to walk it.
How I work
Ships to production
Not prototypes. Everything on this site is live, server-rendered, and running — the same bar I hold for client work.
Honest scoping
If I'm not the right fit, I'll say so up front. You get a straight read on feasibility, effort, and risk before any commitment.
AI where it earns its place
I build with LLMs daily, but only where they beat a simpler approach. No AI theater — just tools that work.
Have something in mind?
Tell me the problem in a sentence or two. I'll give you an honest read on whether I'm the right person for it — usually within a day.