Kelly SETT × AI Strategy

Project Nova

DaySmart Intelligence Briefing
Internal Use Only

Invalid access code. Try again.

Restricted to Kelly SETT team members

KELLY SETT — Project Nova
Confidential — Kelly SETT Internal

Project Nova

How Kelly wins DaySmart's AI transformation — and builds a new model for every pursuit after it.

KELLY SETT × DAYSMART
Scroll to begin↓
The Opportunity

At a Glance

A PE-backed vertical SaaS portfolio with 10 products, 9 acquisitions, and a CTO who wants to cut cycle time in half with AI.

$19.9M
Annual Revenue
~191
Employees (+10.8% YoY)
10
Products in Portfolio
9
Acquisitions Since 2019
6+
Programming Languages
50%
Target Cycle Time Cut
$15K
Phase 1 Discovery
$1-2M+
Full Opportunity (18mo)
Why this matters: DaySmart is PE-backed by LLR Partners ($3.5B+ fund) and Parthenon Capital. They need growth metrics for the board quarterly. Jason Kirk was hired specifically to unify 10 products from 9 acquisitions — and Project Nova is his opening move. If we win Phase 1, we own this relationship for years.
Company Intelligence

The DaySmart Story

From a single desktop salon app to a portfolio of 10 vertical SaaS products — assembled through relentless acquisition.

1999
Founded in Ann Arbor, MI as Salon Iris — desktop software for hair salons. Built on PHP and Windows. The product that started it all.
2016
SFW Capital invests — first private equity backing. Validates the vertical SaaS model and sets the stage for growth.
PE Investment
October 2019
SFW exits → LLR Partners + Parthenon Capital recapitalize. Transaction structured specifically to "increase investment and pursue complementary acquisitions." The roll-up strategy begins.
Growth Recapitalization
December 2020
Acquire AppointmentPlus — enterprise scheduling platform. Brings .NET stack into the portfolio. First acquisition under new PE.
Acquisition #1
February 2021
Patrick Shanahan becomes CEO. Previously COO at CardConnect → grew to 67K+ customers → acquired by Fiserv for $750 million. His mandate: "Be aggressive on the acquisition front."
New CEO
February 2021
Acquire Vetter Software — cloud-native veterinary practice management. Brings Python/Node stack. Already has AI (SOAP notes).
Acquisition #2
July 2021
Acquire Dash Platform — sports/recreation facility management, Seattle. Brings Node.js stack. Third acquisition in 6 months.
Acquisition #3
January 2022
Jason Kirk joins as CTO. Ex-CTO of REPAY (pivotal role in 2019 IPO). Led multiple M&A integrations. Hired specifically to unify the growing platform. 28+ years experience.
New CTO
2022
Acquire TeamUp — fitness/gym management platform, London UK. Brings Ruby on Rails stack + international operations.
Acquisition #4
2023
Two additional acquisitions — expanding the portfolio further. Details to be confirmed in next conversation with Jason.
Acquisitions #5-6
February 2024
Acquire Time To Pet — pet sitting/dog walking software, Austin TX. Brings Ruby on Rails + MySQL (Percona-backed).
Acquisition #7
March 2026
Acquire Slick — UK salon platform, 8,000+ beauty professionals. Expands global footprint. 9th acquisition.
Acquisition #8-9
April 2026 — NOW
Project Nova kicks off. Two teams tasked with building AI infrastructure for CRM integration. They're not allowed to build it manually. The AI transformation begins.
Project Nova
The pattern is clear: This is a PE-backed roll-up. Every acquisition adds customers AND a new tech stack. Jason Kirk was hired to solve the resulting platform chaos. Project Nova is his opening move — and Kelly needs to be right beside him.
Know Your CTO

Jason Kirk

The man driving Project Nova. Understanding how he thinks is the difference between winning and losing this deal.

Jason Kirk
Chief Technology Officer
  • CompanyDaySmart Software
  • SinceJanuary 2022
  • LocationPhoenix, Arizona
  • Experience28+ years in technology
  • EducationBS Management/CIS, Park University
  • RecognitionTop 25 Executives of Arizona, 2024
  • Prior RoleCTO at REPAY (IPO 2019)
  • Before ThatVP Product Dev, CCBill
  • SpecialtyM&A integration, platform scaling, payments

How Jason Thinks — In His Own Words

"You have individuals that are really good at adopting AI and it's making those individuals more effective. But when it needs to be handed off to somebody on their team who isn't leveraging AI — we get no benefit from one person's individual efficiency gains."
He sees the SYSTEM, not the individual. Solutions must work at team level.
"The idea is to have AI do the heavy grunt work, and we become strategists that are helping AI understand the job that it's trying to do."
Not afraid of AI replacing work. He wants humans as orchestrators.
"We have established processes that work and give us good data. It's just — how do we take those processes and hook agents up to them?"
He's NOT asking for process redesign. He wants to AI-enable what already works.
"How can we make these agents like our actual employees that are doing the work, but our employees are overseeing it?"
He thinks of agents as team members, not tools. Speak his language.
"It takes us on average 35 days to get an epic into production. How can we drive that down to 16, 17 days?"
Specific, measurable goals. We should mirror this precision in everything we present.

Jason's Decision Framework

🎯
Customer Impact
Ship features customers want, faster
⚡
Cycle Time
Measurably reduce epic delivery
📐
Scale
Works across ALL 10 products
🔗
Platform
Helps unify the portfolio
🚀
Pragmatic
No studies. Show results fast.
The CEO's Playbook

Patrick Shanahan

The man behind the acquisition machine. His background tells you everything about where DaySmart is headed.

$750M
CardConnect exit to Fiserv (2017)
67K+
Customers he grew CardConnect to
9
Acquisitions under his watch
1-2/yr
Expected acquisition pace

Shanahan was COO at CardConnect, where he grew the company to 67,000+ customers before Fiserv (First Data) acquired it for $750 million in 2017. He knows exactly what a successful PE exit looks like — and he's building DaySmart toward one.

He joined DaySmart's board in 2019 after the LLR/Parthenon recapitalization, became CEO in February 2021, and immediately began executing the acquisition strategy.

"I will work with the leadership team to be aggressive on the acquisition front to deliver solutions that meet the needs of more businesses worldwide."
They WILL keep buying companies. Every acquisition = more tech stack chaos = more work for Kelly.
Our door: Roxy Lee has a personal relationship with Shanahan. This is how we got in. The relationship is warm — they've already met in Phoenix. Shanahan's payments background aligns with Kelly's fintech practice. This isn't a cold call. This is an inside track.
The Technical Challenge

The Tech Stack Problem

"We literally got everything." — Jason Kirk. Ten products. Six languages. Nine acquisitions worth of technical debt.

ProductOriginPrimary StackDatabaseStatus
DaySmart SalonHomegrown 1999PHPSQL Server / MySQLLegacy desktop + cloud
DaySmart PetHomegrownPHP—Legacy
DaySmart Body ArtHomegrownPHP—Mobile + web
DaySmart SpaHomegrownPHP—Legacy
AppointmentsAcquired 2020.NETSQL ServerEnterprise SaaS
Vet (Vetter)Acquired 2021Python NodeCloud-nativeMost modern — has AI
Recreation (Dash)Acquired 2021Node.js—Modern SaaS
TeamUpAcquired 2022Ruby—UK-based
Time To PetAcquired 2024RubyMySQL (Percona)Clean SaaS
SlickAcquired 2026TBD—UK — newest

✅ Standardized Across All

  • Jira — project management
  • Confluence — documentation
  • GitHub — source control (migrating holdouts)
  • AWS — cloud infrastructure
  • Zuora — billing platform
  • Salesforce — CRM

❌ Not Standardized

  • Programming languages (PHP, Python, .NET, Ruby, Node)
  • Frameworks (various per product)
  • Databases (MySQL, SQL Server, others)
  • Deployment pipelines
  • Architecture patterns (monolith vs. micro)
  • API design (different per product)
This Is Our Edge
Most AI consultants optimize for ONE stack. DaySmart needs someone who can make AI work across ALL of them. A stack-agnostic AI playbook that rolls to 10 different teams — that's what we build.
The Initiative

Project Nova

Jason's vision in his own words: turn AI from individual productivity into a team-level system that doubles output.

35 days → 16-17 days
Average epic cycle time today → target with AI agents. A 50% reduction that effectively doubles output.

Jason's Agent Vision — The Flow

Feature Request
→
Product Agent
→
Architecture Agent
→
Developer Agent
→
QA Agent
→
Ephemeral Env
→
Human Review
→
Production

What's Happening Now

  • › 2 teams working on CRM integration
  • › Must build AI infrastructure — can't code manually
  • › Selected their best AI adopters
  • › Using GitHub Copilot + Claude Code
  • › "Rapid experimentation phase" — no platform decision yet
  • › Matt is heading up Nova (needs to be in next meeting)

The Rollout Plan

  • › Prove it on Nova teams (CRM project)
  • › Show metric improvement in cycle time
  • › Roll to next team once proven
  • › Goal: every team by end of 2026
  • › Future: go-to-market + support teams
  • › Jira just announced agent workflows — space is moving fast

Why They Need a Partner — And Why Kelly Lost Last Time

Kelly lost out to an India-based competitor on previous work. Jason told Roxy: "We really wanted you guys to get it, but you really didn't have enough there."

What went wrong: Kelly came with staffing, not strategy. Headcount, not expertise.

What's different now: We come with AI expertise, a real plan, demonstrated understanding of their business, and the specific tools and approaches that match Jason's vision. If we win this, we win back the India work too.

AI Recommendations

The AI Toolkit

Specific, practical tool recommendations for DaySmart's multi-stack development teams. Not theory — tools we can deploy.

Already Using ✅
Claude Code
Anthropic's coding agent. Complex reasoning, multi-file changes across any language.
Already deployed. Best for architecture-level work and multi-file refactoring across their PHP, Python, Ruby, Node stacks.
In Use
GitHub Copilot
Inline code completions, PR summaries, chat in IDE.
Already deployed. Handles day-to-day coding assist across all their languages.
In Use
Recommended for Code 🔧
Cursor IDE
AI-native IDE combining chat, inline editing, and multi-file changes in one interface.
Handles ALL their languages. Composer mode is perfect for the kind of cross-file agent work Jason described.
Recommended
Amazon Q Developer
AWS-integrated AI coding assistant with deep cloud integration.
They're on AWS. Q Developer can optimize infrastructure code and suggest AWS-native solutions across stacks.
Recommended
Agent Orchestration 🤖 — Where the Real Value Is
Claude MCP (Model Context Protocol)
Connect AI directly to Jira, GitHub, and Confluence — the exact tools DaySmart uses across ALL teams.
THIS is what Jason is describing. MCP lets agents read Jira tickets, create PRs, update Confluence — natively. This is how you build Product Agent → Dev Agent → QA Agent.
Critical
Custom Agent Pipeline
Product Agent → Architecture Agent → Dev Agent → QA Agent. Built on Claude's tool use API.
We build this FOR them. Each agent has specific context about their products, processes, and standards. This is the core deliverable.
We Build This
GitHub Actions + AI Gates
Automated CI/CD with AI code review, security scanning, and quality gates.
They're standardizing on GitHub. AI-powered Actions can enforce quality across all 10 products regardless of language.
Recommended
Code Review & Security 🛡️
CodeRabbit
AI code review on every PR. Supports PHP, Python, .NET, Ruby, Node, React, Angular.
Covers ALL their stacks in one tool. Consistent quality bar across every product.
Recommended
Snyk
AI-powered security scanning across every language and framework.
Multi-stack security is critical. One scanner for PHP, .NET, Ruby, Python, Node — and their npm/pip/gem dependencies.
Recommended
SonarQube
Code quality gates with customizable rules per language.
Enforce consistent quality metrics across all 6+ languages. Quality gates in CI/CD pipeline.
Recommended
Security Considerations ⚠️
IP Protection
Configure all AI tools to NOT train on DaySmart's code. Use enterprise tiers with data isolation.
Security
API Key & Secret Management
Establish guidelines for what goes into AI prompts. Never paste credentials, tokens, or PII.
Security
GDPR / Data Residency
UK operations (Slick, TeamUp) require GDPR compliance. AI tools must respect data boundaries.
Security
AI-Generated Code Review
All AI output still needs human security review. Establish AI code review checklist and standards.
Security
The Approach

Kelly's Play

Three phases. Start small, prove value, expand. This isn't a framework — it's a working engagement that produces results.

Phase 1 — The Door Opener
AI Discovery & Readiness
⏱ 2 weeks $15,000
  • Deep-dive into DaySmart's SDLC across all products
  • Map AI readiness per team and tech stack
  • Identify top 3-5 high-impact automation points
  • Assess current tool usage and gaps
  • Interview Jason, Matt, and team leads
Deliverable: AI Transformation Roadmap + Tool Recommendations + Quick Win Playbook
Phase 2 — Prove It
Nova Acceleration
⏱ 6-8 weeks $80-120K
  • Embed Kelly AI engineers with Nova teams
  • Build the agent pipeline (Product → Dev → QA)
  • Integrate agents with Jira/GitHub/Confluence
  • Deploy ephemeral test environments
  • Measure epic cycle time + quality
Success metric: Demonstrate path to 50% cycle time reduction on one team
Phase 3 — The Real Prize
Enterprise Rollout
⏱ 12-18 months $500K-$1M+/yr
  • Roll AI workflows to all 10 product teams
  • Platform unification consulting
  • M&A integration playbook (for future acquisitions)
  • AI product features (compete with Fresha)
  • Ongoing embedded AI talent + advisory
This is the relationship: Multi-year, embedded partnership across every product and every acquisition
$1-2M+ over 18 months
Total addressable opportunity if we execute. Phase 1 is the $15K door opener that unlocks the rest.
Why $15K is justified: "This isn't a study. It's a working assessment that produces an actionable playbook they can use with or without us. We're not asking them to pay for a slide deck — we're asking them to invest in the roadmap that accelerates Project Nova."
"Since you guys are the AI experts — what is Kelly doing with AI internally?"

This question WILL come up. Here's how we handle it:

01 — Be Honest
"We're building our AI practice right now — and that gives us an edge. We're not selling a legacy product or a framework designed five years ago. We're building this capability in real-time with our clients."
02 — Show Proof
"This intelligence briefing you're looking at? Built with AI. Our RaceTrac pursuit portal? Built with AI. We don't just talk about AI workflows — we use them every day to build better outcomes."
03 — Flip It
"The companies ahead on AI adoption hired partners who learned alongside them. We're offering to be that partner — hungry, invested, and working in the same tools your teams use every day."
04 — Land It
"Between our AI capability and Kelly's delivery engine, we make Accenture look dumb. We're not a big firm selling you a framework. We're practitioners who ship."
Action Items

Next Steps

Clear actions to move this from intelligence to execution.

The bottom line: DaySmart likes Kelly. Roxy has the relationship. Jason has a specific problem with a measurable goal. The PE board needs growth metrics. We have the AI expertise to deliver. The only question is whether we show up with enough to prove it. This portal is Step 1 of that proof.