Ops Leadership Briefing — April 2026

SGA AI Platform

What we built in 6 days: production dashboards, AI agent infrastructure, and operational intelligence — all powered by Claude Code.

98 AI Agents
130 Skills
25K+ Lines of Code
6 Days

What We Built

Live products and analysis tools — click any tile to open it.

🟧

Hygiene Department Sprint

Live Dashboard

Full hygiene department analysis — perio percentage benchmarks, code calibration (D1110 vs D4346 vs D4910), production targets, and reactivation strategy across all 260+ practices.

Hygiene KPIs Code Calibration Benchmarks
📦

Procurement Analysis

Live Dashboard

Supply cost analysis, vendor consolidation opportunities, category spend breakdown, and GPO pricing benchmarks for dental supplies across the SGA network.

Spend Analysis Vendor Comparison Savings
👥

Labor Analysis Dashboard

Live Dashboard

Labor cost modeling, staffing ratios, provider productivity, overtime analysis, and compensation benchmarking. Integrates with Power BI data pipeline for real-time updates.

Staffing Ratios Productivity Cost Modeling
🎯

Aspen Dental Research

Competitive Intel

Deep competitive analysis of Aspen Dental — market positioning, practice acquisition model, patient volume strategies, marketing approach, and where SGA has advantage in local markets.

Market Position Strategy Gaps
📊

Office Manager Dashboard

Live Dashboard

Daily operations dashboard for office managers — morning KPI snapshot, production tracking, schedule fill rate, collections, and team performance. Redesigned in V3 with the new reporting engine.

Daily KPIs Production Collections
☀️

OM Morning Brief

Live Report

Automated morning briefing for office managers — yesterday's performance summary, today's schedule overview, open items, and action priorities. Generated daily from the data pipeline.

Auto-Generated Daily Delivery Actionable
🔎

Practice SEO Audits

Live Reports

100-point SEO audit for each practice website — technical SEO, content quality, local search optimization, Google Business Profile, and competitor benchmarks. Scored across 8 categories.

Per-Practice 100-Point Scoring Actionable
🤖

AI Engine Optimization

Live Reports

AEO audits — how each practice appears in ChatGPT, Gemini, and Perplexity AI search results. Citation tracking, knowledge panel coverage, and AI-readiness scoring per practice.

AI Visibility Citations Per-Practice
📈

Power BI Bridge

Infrastructure

Private agent that connects to Power BI via the DAX query bridge — pulls live data from SGA's Azure data warehouse into AI-powered dashboards. Enables automated reporting without manual exports.

Azure → AI Live Queries Private Agent
🎨

Content Engine Platform

Production Code

Full content management platform — 17 screens for content capture, AI processing, editorial review, brand compliance, design studio, and multi-channel distribution across 260+ practices.

25K+ Lines 37 API Files 17 Pages
🌐

Practice Website Prototypes

Live Site

Per-practice website templates with brand-specific colors, photos, and service pages. Each practice gets its own unique site reflecting their local identity — not cookie-cutter corporate.

Per-Practice Brand-Specific SEO-Ready
🛠

Architecture & Wireframes

Full Portal

The complete platform blueprint — architecture explorer, event contracts, 42 interface specs, requirements tracker, PHI data flows, cost models, and 69 wireframe variations.

42 Interfaces 69 Wireframes Interactive

Key Concepts from Last Week's Session

What we covered when we built the SGA agent environment together.

Agents

Specialized AI team members with deep domain expertise. Each agent has a defined role, responsibilities, and knowledge domain — like a virtual staff member who never forgets and never needs onboarding twice.

Example: sga-architect validates decisions against the technical architecture. sga-phi-reviewer catches HIPAA issues before they ship.

Agent Swarms

Multiple agents reviewing or working on the same task simultaneously — different perspectives catching different problems. Like having 5 specialists review your work at once instead of sequentially.

Example: We ran a 5-agent QA swarm on the Content Engine. Each agent found different issues — security, UX, architecture, PHI compliance, performance.

Agent Development

Building new agents from domain knowledge — taking what experts know and encoding it into reusable AI definitions. Created 98 agents covering development, marketing, finance, ops, and SGA-specific domains.

Process: Identify domain → Define role & expertise → Add decision rules → Connect to codebase → Test with real tasks

Skill Development

Skills are reusable knowledge packs that auto-activate when relevant code is being written. They embed best practices so the AI follows them automatically — no need to remember and re-explain every time.

Example: The "drizzle-orm-best-practices" skill auto-activates when writing database queries, ensuring correct schema usage and query patterns.

Knowledge-to-Skill Pipeline

A process for converting existing documents, PDFs, books, and guides into Claude skills. Takes tribal knowledge and makes it permanently available — the organization never loses what it learns.

Example: We converted the SMILE Program docs and dental coding guides into skills that now auto-apply during development.

Wispr Flow

Voice-to-text tool installed on team machines for faster AI interaction. Speak naturally and it transcribes in real-time — dramatically speeds up prompt writing and reduces friction when working with Claude Code.

Instead of typing complex prompts, speak them naturally. Wispr Flow handles transcription with high accuracy, including technical terms.

6-Day Build Log

Day 1 — April 7
Workspace Bootstrap + Wireframes
Created the SGA development workspace. Built 42-interface wireframe gallery with all 5 platform surfaces.
Day 2 — April 8
Architecture + Training Materials
Architecture explorer, event catalog, requirements dashboard, Claude Code training guide with 31-slide deck. Practice Operations wireframes (23 screens, 69 variations).
Day 3 — April 9
Content Engine — Zero to v1.0
Built the entire Content Engine in one day. 9 tech stack skills (32,613 lines). Asset Registry API, Editorial Queue, Studio Portal, Distribution Engine, Dashboard, Capture App. Multi-agent QA review and all fixes.
Day 4 — April 10
Agent Environment Build Session
Added 22 skills, 20 global agents, partner skills & agents library. Rewrote CLAUDE.md with operating principles (self-annealing, confidence check). Set up Wispr Flow. Discussed agent swarms, skill development, knowledge pipelines.
Day 5 — April 11
Marketing + AEO Strategy
Marketing AEO strategist agent. AEO audit reports for practices. AI Engine Optimization scoring and recommendations.
Day 6 — April 12-13
Ops Dashboards + Competitive Intel
Hygiene sprint analysis, procurement dashboard, labor analysis, OM dashboard redesign, morning briefs, Aspen competitive intelligence, Power BI bridge infrastructure.

What This Enables

Scalable Intelligence

Every analysis we build — hygiene benchmarks, procurement savings, labor models — can be regenerated with fresh data at any time. The agents and skills don't forget how to do the work.

Team Multiplication

98 agents means any team member with Claude Code access can tap into deep expertise — dental coding, marketing strategy, financial analysis — without waiting for a specialist's calendar.

Knowledge That Compounds

Every document, every meeting note, every best practice gets encoded into the system. The platform gets smarter every week — unlike traditional tools that stay static after deployment.