Kindling Solutions

What it is

Kindling Solutions is a boutique AI systems and automation agency founded by Jason Katz in Littleton, Colorado. Engagements start with a paid diagnostic called the Blueprint (21 days, $5K) that maps how work actually moves through a client's business and identifies where a human is the real bottleneck, before any build begins. From there, clients choose FullStak, a custom AI agent team installed across the business, or an In-House Agent, a single AI role installed to solve one defined problem. The pitch is explicitly anti-advice: "We don't sell advice. We develop working systems." The agency's signature output has been AI video production — a multi-model pipeline (Higgsfield Soul for character training, GPT Image 2 for anchor frames, ElevenLabs for voice, Claude Code with MCP to orchestrate parallel work) that produces cinematic, character-consistent video for a fraction of traditional agency cost. Cited case-study figures include 10–15 hours a week eliminated per automation, 20% month-over-month growth after a system install, and up to 75% workload reduction for some clients — self-reported, not independently audited. Named case studies include Xerox, Squlpt Body, and Crow Estate Planning. Katz's older domain, katzventures.co, is a separate personal newsletter ("Simplify. Systemize. Scale."), not the consultancy.

Who it's for

Founder-led businesses that want to scale content output, build internal AI agent teams, or produce professional video content without hiring an agency or adding headcount. Particularly strong for founders who have hit a ceiling on DIY AI workflows and need someone to build the documented, repeatable system that turns a one-off experiment into a business capability.

Why it's better

• •Nine months of AI video pipeline refinement distilled into a proven, repeatable system with documented steps.
• •Systems are built to be teachable and future-proof — each capability can eventually become its own product offering.
• •Produces cinematic AI video for ~$300 per piece versus traditional agency quotes in the tens of thousands.
• •Claude Code + MCP orchestration enables 4–5 parallel AI instances coordinating 10–20 agents across separate projects.
• •Transparency-first approach: synthetic for reach, human for trust — no deception baked into the workflow.