Tier 2 · 12 weeks · 16–20 total hours/week, including 6 live hours
LLM Integrations & Automation
Extend a deployed application with AI-assisted data processing and reliable asynchronous workflows. The emphasis is dependable integration, not simply getting a model response.
Who it's for
Developers with 2–4 years of experience moving into applied AI, and students who pass the Tier 1 exit bar or an equivalent assessment.
Entry evidence
Tier 1 exit capability or equivalent: React/API/SQL competence, Git, Docker and a simple deployment — shown through a deployed application, a database/API task and live debugging.
Capability at exit
Add evaluated LLM features and reliable background processing to an application.
What you’ll build and learn
- LLM foundations: tokens, context, embeddings, limitations and task decomposition
- Direct model APIs and a small provider-agnostic interface
- Structured outputs, schema validation and semantic checks
- PDF extraction, OCR and vision trade-offs; document provenance
- Jobs, state transitions, queues and workers
- Server-sent events, retries and idempotency
- Baseline evaluation, logging and cost tracking
- Human review before AI output becomes authoritative
How you’ll show it
- Explain the full processing lifecycle of an AI-assisted job
- Distinguish extraction quality from schema validity
- Handle duplicate work and recover from a failed worker
- Preserve human corrections and show evidence for extracted values
Credential: LLM Application Engineering
Role direction: AI-enabled full-stack/backend engineering or junior applied-AI work, depending on demonstrated depth and prior experience.
Role readiness is assessed; it is not conferred by attendance.