David George SmithSenior Full Stack Engineer · Engineering Manager · Product Owner · AI Enablement Lead

AI Orchestrator & Technical Product Lead

I design the goals. Agents do the labor. I govern the system.

I build the AI-powered systems engineering teams run on — MCP integrations, LLM automation pipelines, Copilot adoption dashboards — and lead the people who use them. I sit at the layer above the automation: I set strategy, manage engineers, own product decisions, and measure what AI is actually worth.

See the work Interview my agent

02 / License optimization identified

25%

Copilot seat usage analysis surfaced a 25% license optimization opportunity.

03 / Operational cost reduction

15–35%

Documented 15–35% operational cost reductions across engineering teams using AI tooling.

04 / Internal agent tools shipped

10+

Custom internal agent tools deployed by engineers trained through the AI KT Series.

05 / Less developer rote work

40%

Reduction in repetitive developer work from MCP-based agent tooling.

06 / ROI model(value − investment) ÷ investment

What is AI actually worth to your team?

Annual ROI

+907%

Labor savings: $874,000
Productivity lift: $0
Investment: −$86,800
Net value: $787,200

ROI = (value − investment) ÷ investment × 100. Value = hours saved × hourly cost × 46 wks + lift. Illustrative inputs — set them to your org.

07 / Ask my resume local-first + mcp

Ask anything about Dave's work. Answers are grounded in his profile through MCP tools and routed local-first: a model on Dave's own AI stack, with Claude as the paid fallback.

08 / Bring it to your own agent

Interview me from your own Claude.

The assistant on the left runs on a public MCP server. Add it to Claude Code (or any MCP client) and ask about my experience, metrics and case studies in your own tools.

claude mcp add --transport http dave-smith https://smithdavedesign.herokuapp.com/v4/mcp

Exposed tools

ƒ get_experience_summary
ƒ retrieve_roi_metrics
ƒ list_case_studies
ƒ get_case_study
ƒ list_builds
ƒ get_build
ƒ get_stack
ƒ get_contact

P1 / pillar

Orchestration Mastery

Moving teams from building to spec to designing goals for autonomous agents — and choreographing multiple specialized agents into one reliable workflow.

P2 / pillar

Ethical Stewardship

A Coding for Good background applied to agents: defining how they are gated, monitored, and aligned with human values.

P3 / pillar

The Force Multiplier

Training a technical workforce to build its own agentic tools, so AI capability scales past any one person.

High-ROI orchestration

Action · Tooling · Metric

case / A

The Governance & ROI Engine

Optimizing enterprise intelligence via Copilot analytics.

Action: Engineered a real-time dashboard for Copilot usage reporting and analysis — benchmarks, adoption trends, team-level comparisons and an ROI calculator built for CTOs and engineering managers.
Tooling: Telemetry ingestion · Custom visualization layer · TypeScript · React · ROI model
Metric: Identified a 25% license optimization opportunity and documented 15–35% operational cost reductions across engineering teams.

case / B

Building the Digital Workforce

Scaling agentic capability via MCP training — the AI KT Series.

Action: Architected an org-wide Knowledge Transfer series training engineers to build Model Context Protocol servers, Skills, and custom instructions — the human infrastructure that makes AI adoption stick.
Tooling: Model Context Protocol · Claude Code · GitHub Copilot · Custom Instructions · Skills
Metric: Accelerated the deployment of 10+ custom internal agent tools, reducing developer rote work by 40%.

case / C

AI-Powered Developer Tooling

Automating the SDLC via intelligent code agents.

Action: Developed a suite of AI-integrated coding tools that streamline repository management and pull-request review.
Tooling: LLM agents · Map-Reduce-Align · GitHub API · Node.js · MCP
Metric: Shorter review cycles and consistent PR quality through structured, evidence-backed agent recommendations.

case / D

Figma → Jira / Wiki Automation

Eliminates manual documentation labor between design and engineering across the org.

figma_mcp
llm_draft
Human checkpoint: pm_review
jira_wiki

case / E

QA Test Case Automation

Reduced manual QA effort while keeping quality standards consistent across teams.

storybook
mcp
casegen
Human checkpoint: qa_review

Things I've built

screenshots from the real apps

build / 01live

RepoHQ

Portfolio intelligence for every repo I own — and an agent factory that acts on it.

Next.js 16 · TypeScript · Neon Postgres · Drizzle ORM · Auth.js (GitHub OAuth)

build / 02live

AI Radar

Weekly, signal-based rankings for 150 AI tools — no opinions, just data.

Next.js 16 · TypeScript · Neon Postgres · Drizzle ORM · Inngest

build / 03live

Yosemite National Park destination page with a Wikimedia hero photo of Half Dome, difficulty, trip length, budget, best months and the NPS-sourced entrance fee

Open Adventure

“Skyscanner Explore” for the outdoors — discover where to go, not just research a place you picked.

Next.js 16 · TypeScript · Postgres + PostGIS · Prisma · MapLibre / MapTiler

build / 04live

Open Fly trips dashboard with trip cards, budgets, collaborators and quick tools for expenses, chat, flights and visas

Open Fly Travel Assistant

Group trip planning where AI reads the booking confirmations for you.

Next.js · TypeScript · Supabase · Claude API · Mapbox

build / 05live

Roots and Branches tree view showing Anakin Skywalker and Padmé Amidala connected to their children Luke and Leia

Roots & Branches

An interactive family tree you can explore, enrich and share.

React + Vite · Node.js / Express · Supabase Postgres · React Flow · Stripe

09 / Force multiplier

AI KT Series: Scaling the Agentic Workforce

Tools don't scale an org — people who can build tools do. The KT series turned engineers into MCP builders, with governance designed in from module one.

01MCP fundamentalsHow models reach tools and data; when to build a server versus a skill.
02Building toolsEngineers ship an MCP server against a real internal system.
03Skills & IDE optimizationClaude Code, Copilot and custom instructions tuned per repo.
04Gating & HITLCustom instructions that scope agents; mandatory human checkpoints.
05Agentic workflowsMulti-step agent runs with clear owners and stop conditions.

Outcome: 10+ internal agent tools shipped · 40% less rote developer work

Ethical stewardship

Coding for Good → agent governance

Autonomy is earned, scoped and reversible. Every agent I deploy sits at an explicit level on this ladder, with a named human above it.

L0
Suggest
Agent drafts; human writes. Default for anything customer-facing.
HITL: Every output
L1
Draft & route
Agent produces an artifact and routes it to a named reviewer (tickets, test cases, PR reports).
HITL: Before publish
L2
Act in sandbox
Agent executes inside a scoped environment with gated skills and logged tool calls.
HITL: Before merge / deploy
L3
Act with hard stops
Agent runs end to end; defined hard stops (security, money, people data) always escalate.
HITL: On hard stop

Stack, and why

hover or focus for rationale

I design the goals. Agents do the labor. I govern the system.

What is AI actually worth to your team?