Available
Available for high-leverage work
Software engineering · AI systems

SOFTWAREENGINEER&AISYSTEMS.

I build products that make complex systems legible — realtime apps, AI-integrated workflows, and information surfaces that turn full-stack engineering depth into a visible advantage.

Latest ranked signal
Introducing Cache Response Rules
CLOUDFLARE11m agoscore 60Open signal feed →
Experience
10+ yrs
Live sources
14
Shipped systems
6
Published essays
2
Featured builds
5
Signals ranked
28

A portfolio that behaves like operational software.

The strongest version of this site is not louder. It is more instrumented: live data, authenticated publishing, ranked inputs, and proof surfaces that make Gordon Todd Watts feel like the engineer behind a system, not just the owner of a landing page.

Live sources
14
Ranked items
28
Freshest signal
<2h
Categories
6
Published posts
2
Publish API
locked
Product engineering
Full-stack systems

Interfaces, APIs, content pipelines, and operational workflows shaped as one product surface.

AI integration
Agent-ready workflows

MCP-aware publishing, AI product strategy, and LLM engineering treated as infrastructure, not garnish.

Realtime posture
Signal-driven UX

Live feeds, ranked data, and status surfaces that show technical awareness in the interface itself.

Where Gordon creates leverage.

The strongest fit is work where product taste, full-stack execution, AI systems, and operational clarity all matter at once. Gordon can move from fuzzy strategy to shipped software without losing the thread between interface, API, data, deployment, and proof.

Case files7
Live systems6
AI-linked3
Quality gates6
Capability ownership mapFive lanes, one accountable engineer
Inspect full capability map →
01
Capability laneProduct engineering

Turns ambiguous business goals into shippable product systems.

Next.jsReactTypeScript
Interfaces, workflows, API contracts
02
Capability laneAI application systems

Builds LLM-backed workflows where model output has product context and control surfaces.

LLM APIsagentsRAG patterns
AI product layers that behave like infrastructure
03
Capability laneBackend and data flow

Keeps APIs small, inspectable, and operationally visible.

route handlersRSS/API aggregationMDX
Dashboards, source health, durable event trails
04
Capability laneCloud and operations

Treats deployment, monitoring, and security posture as part of the product.

VercelAzureCI gates
Production-ready posture before handoff
05
Capability laneStrategic systems thinking

Uses marketing leadership experience to build software that explains its value quickly.

positioningconversion pathsanalytics
Technical products with clear commercial signal

The release bar is part of the interface.

Typed contracts, source telemetry, browser semantics, security headers, and performance budgets are checked locally before handoff. The detailed evidence remains inspectable through Ops and the portfolio API.

Local release barbun run verify
1,908 checks // 0 failures
Typecheckpassing
bun run typecheck

TypeScript contract across app routes, components, and shared data models.

Enterprise audit503 checks
bun run audit:enterprise

Static checks for positioning, design tokens, source quality, semantics, security posture, and proof surfaces.

Build artifactpassing
bun run build

Next.js production build proves route boundaries, generated assets, and dynamic API surfaces.

API contract823 checks
bun run audit:api

Built route handlers prove OpenAPI, profile proof, feed telemetry, health, ops, bounded publishing, rate policy, atomic writes, and monitor shape.

Browser audit396 checks
bun run audit:browser

Desktop/mobile route smoke, generated images, structured data, security headers, axe, overflow, keyboard flow, sanitized Blog output, and semantic controls.

Performance audit186 checks
bun run audit:performance

Core Web Vitals, interaction response, Blog budgets, transfer budgets, visible content, luminance diversity, and Ghost Signal accent pixels.

✦ Selected Work

Systems I've Built

View all case files →
JPope Tax Portal
LIVE
2024

JPope Tax Portal

Enterprise client portal for tax consultancy

Full-featured client portal built for JPope Tax Consultancy. React, TypeScript, Next.js frontend backed by Convex for real-time data sync. Role-based access, document management, and client communication tools.

ReactTypeScriptNext.jsConvexEnterprise
Surface
Portal + docs + comms
Mode
Realtime-ready
Focus
Security + clarity
Outcomes
Unified client communication surface
Architecture
Role-based accessRealtime data syncDocument workflows
JPOPE Tax Consultancy
LIVE
2026

JPOPE Tax Consultancy

JamiePope.com commercial real estate tax website

Public website for JPOPE Tax Consultancy that turns a complex CRE tax advisory practice into a navigable planning system: service taxonomy, lifecycle routing, resources, newsroom content, and conversion paths for owner and advisor inquiries.

ReactTypeScriptInformation ArchitectureSEOCRE Tax
Experience
30+ yrs
Service areas
23
Planning lanes
6
Outcomes
Makes specialized CRE tax work easier to navigate
Architecture
Service-route taxonomyLifecycle planning frameworkResource + newsroom content model
TaxScout.app
LIVE
2025

TaxScout.app

AI-powered commercial real estate tax optimization

Commercial real estate tax optimization platform for multifamily properties. Cost segregation analysis, 1031 exchange calculations, REPS qualification, and accelerated depreciation schedules. MCP-connected for AI agent access.

Next.jsTypeScriptAI/MCPReal EstateTax Engineering
Domain
CRE tax workflows
AI layer
MCP-connected
Product move
Decision support
Outcomes
Turns niche tax logic into a usable product
Architecture
AI agent publishing surfaceTax calculation workflowsMultifamily domain modeling
MAINFRAME
LIVE
2025

MAINFRAME

Open-source bash stdlib for AI agents

4,000+ pure bash functions: JSON, strings, arrays, HTTP, CSV, git, Agent Working Memory (AWM), RAG, embeddings, and multi-agent orchestration. The codex for AI-native shell scripting.

BashOpen SourceAI AgentsAWMRAG
Functions
4,000+
Runtime
Pure bash
Use case
AI agent systems
Outcomes
Signals systems-level engineering range
Architecture
Pure bash runtimeAgent working memory primitivesCLI-first orchestration surface
PropertyJunk.com
LIVE
2024

PropertyJunk.com

Junk removal SaaS with integrated CRM

Complete SaaS platform for junk removal operations. Real-time fleet tracking, employee management, customer CRM, scheduling, and billing—all in a single unified interface. Built for scale.

Next.jsTypeScriptConvexReal-timeSaaS
Mode
Ops SaaS
Surface
CRM + fleet + billing
Focus
Workflow continuity
Outcomes
Turns field operations into software
Architecture
Operations dashboardFleet and scheduling workflowsUnified CRM + billing
✦ /now

What I'm Doing
Right Now

Building
ACTIVE
PropertyJunk.com — junk removal SaaS platform
ACTIVE
TaxScout.app — commercial real estate tax optimization
WIP
MAINFRAME — open source bash stdlib expansion
WIP
toddwatts.dev — this site
Reading
01Attention Is All You Need (re-read)
02The Pragmatic Programmer — 2nd ed.
03Designing ML Systems — Chip Huyen
Watching
Andrej Karpathy
Neural nets: Zero to Hero
Lex Fridman
Sam Altman ep.
ThePrimeagen
Rust streams
My Setup
EditorVS Code + Claude Code
AIClaude Sonnet 4.5, Pi
RuntimeBun 1.2
DeployVercel
Gordon's workspace