Skip to main content
arrow_backAll Courses
BEST FOR: FOUNDATIONS GRADUATES

Intermediate AI Product Development

Tier 02

computerVirtualcalendar_monthStarts May 21, 2026scheduleThursday · 7:00 PM EDT

Course Overview

This program bridges the gap between building AI demos and delivering AI products that serve real users and real businesses. Participants work with external stakeholders, handle ambiguous requirements, manage production infrastructure, and deliver software that someone other than themselves depends on. This is not a tutorial program. This is client-facing, revenue-capable product engineering.

Goal Outcome

Deliver a client-facing, production-grade AI product to a real stakeholder.

Who This Is For

Graduates of the Foundations of AI-First Product Development program ready to work with real clients.

$2,500

Program Fee

Prerequisites

Completion of Foundations of AI-First Product Development program.

Core Tools

Next.js, Supabase, OpenAI/Anthropic, Vercel, GitHub, Stripe, Sentry, Resend

Start Date

May 21, 2026 · 8 weeks

Your Instructors

Jason A. Oglesby

Jason A. Oglesby

Founder, Ergon Insights

Jason A. Oglesby is a technology leader and AI transformation strategist with 30+ years of software development experience. As the former CTO of Rev.io, a leading SaaS company, he led a 70+ person IT organization through a full-scale AI transformation where they created more software over his last 12 months than they had completed in the previous decade, all by embedding AI-first practices across every layer of the engineering process. Now running Ergon Insights, Jason consults with organizations on practical AI adoption, coaches technology leaders, and builds AI-powered software solutions using AI-first development practices and tools as core accelerators. Jason holds an MS in Business Analytics from the University of Nebraska-Lincoln and brings deep experience from leadership roles at IQVIA, Healthcare Data Solutions, and other technology companies. A veteran of the Army National Guard and a leadership practitioner grounded in Extreme Ownership principles, he's passionate about building people up alongside building technology. Based in Johnson City, Tennessee, Jason is committed to growing the Tri-Cities region into a technology hub, making Forge Academy a natural extension of that mission: equipping everyday problem-solvers with AI tools to build real software solutions in just six weeks.

Course Curriculum

1
Week 1Real Requirements & Stakeholder Discovery
expand_more

Participants learn that building the right thing matters more than building things right. "Real users don't give you a spec — I have to extract it."

Topics

  • ›Translating business problems into product requirements
  • ›Stakeholder interview techniques
  • ›Writing user stories from conversations, not assumptions
  • ›Scope negotiation: what to build first vs. what to defer
  • ›Introduction to the semester-long client engagement model

Deliverables

  • checkRecord a 30-minute structured stakeholder interview (video or audio)
  • checkWrite a 1-page Problem Statement Document
  • checkProduce a prioritized feature list with MoSCoW categorization
  • checkCreate a user persona document based on actual interview data
  • checkWrite reflection: what the stakeholder actually needs vs. what they asked for
2
Week 2Multi-Model AI Architecture
expand_more

Participants understand AI architecture is an engineering decision, not a default. "I choose the AI — it doesn't choose me."

Topics

  • ›Choosing the right model for the right task
  • ›Prompt chaining and multi-step AI workflows
  • ›Structured output parsing with validation layers
  • ›Fallback strategies: what happens when the AI fails
  • ›Cost optimization across model tiers (GPT-4o vs. GPT-4o-mini vs. Claude Haiku)

Deliverables

  • checkBuild and deploy a multi-step AI content pipeline for the stakeholder's domain
  • checkImplement model selection logic routing simple tasks to cheaper models
  • checkAdd structured JSON output with schema validation and fallback handling
  • checkBuild cost tracking per generation (tokens used and estimated cost)
  • checkSubmit architecture diagram showing model routing logic
  • checkWrite cost comparison report: same task across 3 model tiers
3
Week 3Database Design & Data Modeling for AI Products
expand_more

Participants build data foundations that can support production AI products. "The database is the product — the UI is just the window."

Topics

  • ›Relational data modeling for multi-tenant applications
  • ›Designing schemas that support AI features (storing prompts, outputs, feedback)
  • ›Row-level security policies for real multi-user environments
  • ›Database migrations and schema versioning
  • ›Indexing strategies for search and retrieval

Deliverables

  • checkDesign and implement a multi-table relational schema (minimum 4 tables)
  • checkImplement row-level security enforcing user/organization data isolation
  • checkBuild an AI interaction logging table (prompt, response, model, tokens, latency, feedback)
  • checkImplement full-text search using Supabase pg_trgm or tsvector
  • checkWrite a seed data script for development and demo purposes
  • checkSubmit entity-relationship diagram, versioned SQL migration files, and RLS policy documentation
4
Week 4Payments, Metering & SaaS Economics
expand_more

Participants understand that a product without a business model is a hobby. "I can build something people will pay for."

Topics

  • ›Stripe integration fundamentals (test mode)
  • ›Subscription models vs. usage-based pricing vs. credit systems
  • ›Metering AI usage per customer
  • ›Tiered access control (free, pro, enterprise)
  • ›Unit economics: cost per AI call vs. revenue per customer

Deliverables

  • checkIntegrate Stripe Checkout (test mode) with subscription tiers
  • checkBuild usage metering to track AI calls per user per billing period
  • checkImplement tiered access gates (free tier: limited calls/day; paid: unlimited)
  • checkDeploy billing dashboard showing current usage and plan details
  • checkCreate pricing page with tier comparison
  • checkComplete unit economics worksheet: cost to serve 100 users at each tier
  • checkWrite reflection on pricing strategy rationale for the client project
5
Week 5Production Hardening & Observability
expand_more

Participants operate production infrastructure, not demo environments. "If it's not monitored, it's not in production."

Topics

  • ›Error monitoring with Sentry
  • ›Structured logging for AI applications
  • ›Performance monitoring and latency tracking
  • ›Rate limiting and abuse prevention
  • ›Environment management (dev/staging/production)

Deliverables

  • checkIntegrate Sentry error tracking capturing real errors
  • checkImplement structured logging for all AI interactions (model, tokens, latency, success/failure)
  • checkAdd rate limiting on AI endpoints (per user and per IP)
  • checkCreate a health check endpoint returning system status
  • checkConfigure environment-specific settings (dev vs. production API keys, feature flags)
  • checkSubmit Sentry dashboard screenshot, log sample, rate limiting test results, and completed Production Readiness Checklist
6
Week 6External Integrations & Workflow Automation
expand_more

Participants build systems that live inside real business workflows, not in isolation. "My product connects to the way people already work."

Topics

  • ›Webhook design and implementation
  • ›Third-party API integration patterns
  • ›Email notifications and transactional messaging
  • ›Background jobs and scheduled tasks
  • ›Building for the stakeholder's existing tool ecosystem

Deliverables

  • checkBuild a webhook endpoint receiving data from an external source
  • checkImplement outbound email notifications via Resend (e.g., weekly AI-generated summary)
  • checkAdd a scheduled background task (e.g., nightly data processing or report generation)
  • checkIntegrate with one stakeholder-relevant tool (Google Sheets, Slack, CRM, etc.)
  • checkImplement retry logic for failed external API calls
  • checkSubmit integration architecture diagram and written error-handling strategy for each integration
7
Week 7Client Delivery Sprint
expand_more

Participants experience the discipline of delivering to someone who is counting on them. "Shipping to a real person changes everything."

Topics

  • ›Feature freeze discipline
  • ›User acceptance testing with real stakeholders
  • ›Bug triage and priority management
  • ›Handoff documentation
  • ›Managing stakeholder feedback without scope creep

Deliverables

  • checkFeature freeze by Wednesday — no new functionality after midweek
  • checkConduct and record stakeholder walkthrough session by Thursday
  • checkResolve all critical bugs before Friday
  • checkWrite technical handoff document (architecture, env vars, schema, known limitations)
  • checkProduce a non-technical user guide for the stakeholder
  • checkSubmit stakeholder feedback summary and bug report with resolution log
8
Week 8Demo Day & Portfolio Presentation
expand_more

Failure to deploy = incomplete. Stakeholder must confirm the product was delivered. "This is what it means to ship."

Topics

  • ›Live deployed application demonstration
  • ›Problem → Solution narrative with stakeholder context
  • ›Architecture walkthrough: data model, AI pipeline, integrations
  • ›Business model and unit economics
  • ›Stakeholder testimonial or written feedback
  • ›Performance metrics: uptime, usage, cost, error rate
  • ›Lessons learned and what you would do differently
  • ›90-second elevator pitch: why this product matters

Deliverables

  • checkPresent live deployed application — failure to deploy = incomplete
  • checkDeliver Problem → Solution narrative with stakeholder context
  • checkWalk through architecture, AI pipeline, and integrations
  • checkPresent business model and unit economics
  • checkObtain stakeholder testimonial or written feedback confirming delivery
  • checkSubmit performance metrics report (uptime, usage, cost, error rate)

Ready to Build?

Join the inaugural cohort and start shipping AI-powered products from week one.

View All Courses
Intermediate AI Product Development | FoundersForge AI Academy