Standex Digital Logo
Engineering Curriculum 2026

Foundations & Modern LLMs

Master the internals of modern LLMs. Understand how GPT-4o, Claude, and Gemini actually process tokens and how to architect systems around them using advanced prompting and structured outputs.

Syllabus Architecture
1
Session 01 IntensiveLive Lab

Session 1 — Modern AI System Internals

  • Evolution of LLMs: From transformers to GPT-4o, Claude 3.5, and Gemini 1.5 Pro
  • How LLMs actually work: tokenisation, temperature, sampling, top-p, stop sequences
  • Context windows explained: what goes in, what gets lost (lost-in-the-middle phenomena)
  • Comparing model capabilities: benchmarks vs real-world production performance
  • AI system design patterns: direct API, wrapper service, and gateway proxies
  • Latency budgets and cost modelling: calculating TCO per token across providers
  • Theoretical Audit: Reading AI vendor documentation like a senior engineer
2
Session 02 IntensiveLive Lab

Session 2 — Advanced Prompting & Structured Logic

  • Prompt engineering fundamentals: role prompting, few-shot examples, and chain-of-thought
  • System prompts vs User prompts: defining 'unbreakable' personality and constraints
  • Structured JSON outputs: response_format, strict mode, and Zod schema validation
  • Prompt chaining: breaking complex tasks into reliable, sequential AI steps
  • Common failure modes: hallucination, context stuffing, and prompt injection security
  • Hands-on: building a reliable schema generator for unstructured data

Module Toolchain

Next.js
Next.js
OpenAI
OpenAI
Vercel
Vercel
GitHub
GitHub
Cohort 04 — Active Enrollment
Early Access Pricing

$1800

$3500
Duration6 Weeks
Next CohortMay 5
Sessions2h 30m / session
Enroll Now
Verified Applied Cohort 04 Graduation Path

Trusted by Engineering Teams

Verification Phase

Capstone Graduation Mastery

Demonstrate your engineering expertise to the world.

Mastery Requirements

The capstone requirement is a **full AI product** with all the features taught in the course—autonomous agents, RAG, usage-based billing, and deep observability. Every project is **personally tested by our team** for production-grade reliability and architectural integrity.

Multi-Agent Orchestration
Vector-driven Memory (RAG)
Stripe Token-Billing
Azure/Vercel Auto-Scaling
Observability & Analytics
Production README & Case Study
SXAI-V26-04

Mastery Level 04

Applied AI Engineer