The Stateless Founder

If you're stuck between $8K-$20K MRR running multiple offers across different business models, this episode breaks down the real numbers behind three paths for nomad founders: AI-augmented services, productized services, and PLG micro-SaaS. Santi and Kira analyze actual case studies from PhotoAI ($100K+ MRR at 87% margins), DesignJoy ($130K MRR solo), and automation agencies to show you the margin profiles, channel requirements, and capacity constraints of each path. Plus, they introduce the Leverage Map - a 90-day focus template with weekly KPIs, cost calculators, and nomad-specific risk planning to help you pick one path and execute it before your next visa run forces the decision.

Before you scale AI content delivery, install a three-layer QA Wall that catches bad outputs with golden-set tests, lightweight LLM judges, and strategic human sampling—plus spend guards that prevent runaway costs while you're offline. Kira shares how a fabricated citation that reached 11,000 LinkedIn followers taught her to build redundancy at every level, from automated heuristics that catch PII and broken JSON to dual-judge systems that escalate disagreements to humans. You'll get the exact golden-set template, judge rubric, sampling math, and Slack alert payloads that keep clients happy and your margins intact.

Escape the custom work trap that's killing your margins and location independence. Santi and Kira break down the exact 90-day plan to transform your AI service into a productized offer with strict boundaries, predictable margins, and a clear path to micro-SaaS spinouts. Learn from real case studies including Design Pickle's $1M run rate growth, Testimonial Hero's $3.4M automation-powered business, and Audience Ops' high-six-figure exit. Get the complete framework: offer one-pager template, weekly KPI dashboard, de-risk checklist with AI model cost ceilings, and the moat scorecard that passes the Lisbon Test. No more 2 AM client calls or scope creep eating your profits—just a systematic approach to building a location-independent business that actually scales.

Turn cold outreach from volume-first to relevance-first by pairing a lightweight AI lead-research agent with a dual-channel schedule (LinkedIn + email), strict daily caps, and reply classifiers that auto-triage interest while protecting domains and inbox placement. Santi and Kira break down the compliance-first system that keeps your infrastructure safe while booking more calls from fewer, better-targeted touches.

With 100 days until the EU AI Act's general application date, nomad agencies and micro-SaaS deployers need a practical compliance plan that doesn't stall growth. Santi and Kira break down the MVCP — Minimum Viable Compliance Plan — a 30-day checklist covering AI disclosure pages, evidence logs, incident playbooks, and vendor tracking. They explain what actually applies to small deployers versus providers, how to price compliance overhead into retainers, and what you can safely ignore for now. Plus: how to turn compliance into a revenue stream by offering "compliance-in-a-box" services to other nomad businesses.

Every travel day without an offline AI stack costs you deadlines and forces you to pay for cloud APIs you can't reach. This episode builds the fix: a portable system that runs summarization, transcription, and tagging locally on your laptop, queues everything to SQLite, and syncs when you're back online. Santi and Kira walk through the complete architecture - from LM Studio and whisper.cpp to battery management profiles and sync conflict policies. You'll learn which tasks to run locally versus defer to the cloud, how to set up file watchers and background workers, and why SQLite with Litestream creates the most reliable offline-first workflow. Plus: the exact hardware requirements, encryption setup, and weekend implementation plan that turns any laptop into a self-contained AI workstation.

Three outages across two providers in 34 days exposed how fragile single-vendor AI stacks really are. When Claude went down on April 6th and 7th, and OpenAI had elevated errors in March, nomad founders running client delivery had dead air - no fallback, no queue, just silence while clients waited for deliverables nine time zones away. This episode breaks down a three-tier failover system you can ship in a weekend: Hot (same provider, different model), Warm (different provider entirely), and Cold (graceful degradation with human-in-the-loop). Plus budget guardrails tied to Google's new Gemini spend caps so your API bill doesn't bankrupt you while the failover does its job. Santi walks through the actual architecture - circuit breakers, jittered retries, Redis queues, and idempotency keys - while Kira pressure-tests the human edge cases and client communication templates you need before the next outage hits.

Nomads lose thousands in leads while sleeping across time zones. This episode walks you through building a five-intent AI voice concierge that books calls, qualifies prospects, and handles pricing deflection 24/7 — with strict spend guards and escalation logic so you never wake up to surprise bills. Santi breaks down the exact per-minute costs of OpenAI's Realtime API versus split STT/TTS stacks, while Kira pressure-tests the human handoff moments and consent scripts. Includes the complete starter pack: call-flow JSON, Make/Zapier blueprints, spend-guard scripts, and a monitoring dashboard template.

Cut through the 2026 "AI agent" hype with three production-ready blueprints that actually survive real client conditions. Santi and Kira break down lead qualification, weekly reporting, and inbox triage agents with exact tech stacks, error handling, monitoring, and pricing math. Learn why bounded agents with human-in-the-loop outperform autonomous systems, plus get the complete implementation sprint to ship your first agent in 14 days. Includes JSON schemas, acceptance tests, and the "Lisbon Test" checklist for location-independent reliability.

If you're running an AI wrapper or automation service without real moats, you're one model update away from losing half your MRR. This episode breaks down five practical defensibility strategies nomad founders can build without a data team: distribution control, workflow lock-in, small-data capture loops, switching costs, and SLA commitments. Santi and Kira walk through real examples from TypingMind, HeadshotPro, and Superhuman, then score different wrapper archetypes using their Defensibility Scorecard. You'll get a 14-day sprint plan to add at least 20 points to your defensibility score, plus templates for HubSpot/Slack integrations and SLA snippets. The episode includes the "Lisbon Test" for reliability and addresses the counterargument about model parity making all wrappers obsolete.

Most nomad founders fail at client delivery not because of skill, but because their operations aren't built for asynchronous, multi-timezone work. This episode ships a battle-tested AI + human system that keeps projects moving while you're asleep: intake forms that auto-scope work, timezone-aware task routing, AI-powered QA checks, and weekly status packs that send themselves at 9 AM client time. Santi and Kira break down two complete stacks (low-code and code-first) with exact pricing, failure modes, and the prompts for AI status generation. Plus the fallback plan for when your OAuth tokens expire mid-flight and you're trekking in Nepal.

Build a WhatsApp AI concierge that captures leads, qualifies them with four targeted questions, and books paid calls automatically — running 24/7 while you sleep, travel, or focus on delivery. We break down the exact costs per conversation (including Meta's 2026 pricing changes), the Make.com blueprint to build it, and the compliance guardrails that keep you from getting throttled. From Instagram ads to booked calls in under 10 minutes, with real unit economics showing 500X+ returns when structured correctly.

If you're running client automations on Zapier, Make, or n8n without proper error handling, monitoring, and retry logic, you are one API change away from losing your biggest client. When you're twelve time zones away, you won't even know it happened. Santi learned this the hard way when he lost an $8,000/month client while island-hopping in Thailand. A Make.com API change broke his automation for four days straight — silent failure, no alerts, forty-eight leads vanishing into the void while he sat on a beach thinking everything was fine. Today we're fixing that. By the end of this episode, you'll have the exact setup to catch failures, route alerts, and auto-retry the stuff that can be retried — all while you're offline. We cover platform-specific error handling for Zapier (custom error handlers + Zapier Manager alerts), Make (incomplete executions + auto-retry), and n8n (global error workflows + per-node settings). Plus retry patterns, incident classification (P0/P1/P2), and the security wake-up call from January's n8n "Ni8mare" vulnerability that exposed 59,000 instances. Downloads include: Notion incident runbook template, CSV error taxonomy, and copy-paste alert recipes for all three platforms.

If you're running five or more AI service clients and still pricing like a SaaS subscription, you're leaving thousands on the table every month. This episode breaks down exactly how to repackage your existing AI agents and workflows into a three-tier structure that clients understand and will pay real money for. Santi and Kira walk through a complete case study showing how to transform a $199/month lead follow-up bot into a $2,500 setup plus $750-$1,500 monthly retainer by adding three simple elements that cost almost nothing but completely change value perception. You'll learn the pricing levers that matter (industry multipliers, SLA windows, compliance premiums), how to handle the "we can build this in-house" objection, and why token costs being dirt cheap actually strengthens your pricing power. Plus: the Lisbon Test checklist for building truly location-independent AI services and a downloadable pricing calculator to find your real margins.

Santi lost his biggest client ($8K/month) while island-hopping in Thailand when a Make.com API update broke his lead routing system. Four days of silence later, the client was shopping for alternatives. This episode maps the seven core systems every location-independent AI business needs to run async across time zones: Intake, CRM/Lead Nurture, Project/Client Ops, Knowledge/Docs, Content Engine, Finance Ops, and Monitoring & Alerts. We break down the unit economics across Zapier, Make, n8n, and Pipedream, show you how to implement retry logic with exponential backoff, and explain why idempotency keys and dead-letter queues aren't just fancy concepts—they're the difference between a business that runs itself and one that breaks the moment you lose wifi. Includes real case studies from Slate Magazine (2,000+ leads via Zapier Agents), Laudable ($240K saved with 200+ automations), and System AI (4-minute tasks cut to 20 seconds). Plus a 30-day challenge to audit your stack and the calculator to model costs across platforms.