AI-102 → AI-103: Azure AI Apps and Agents Developer Associate
Microsoft retiró el examen AI-102 el 30-jun-2026. Su sucesor es AI-103 — Azure AI Apps and Agents Developer Associate, orientado a construir apps de IA generativa y soluciones multi-agente con Microsoft Foundry usando Python.
- Página oficial: Azure AI Apps and Agents Developer Associate
- Study guide: AI-103 skills measured
- Contexto verificado: ver Fuentes y Verificación v4
Verifica el estado (beta/GA) y los skills measured en la página oficial antes de calendarizar.
⚠️ Este track se está actualizando de AI-102 a AI-103. Los retos se migrarán al nuevo alcance.
Exam Overview (AI-103)
- Exam: AI-103 (sucesor de AI-102, retirado 30-jun-2026)
- Level: Associate
- Target Audience: AI Engineers, Solution Architects
- Enfoque: apps de IA generativa + soluciones multi-agente con Microsoft Foundry (Python)
- Passing Score: 700/1000
- Validity: verificar en la página oficial
Los datos históricos de AI-102 (redesign de abril 2025) se conservan abajo como referencia hasta completar la migración a AI-103.
Referencia histórica — AI-102 (retirado)
- Exam: AI-102 (Redesigned April 2025)
- Level: Associate
- Target Audience: AI Engineers, Solution Architects
- Cost: $165 USD
- Duration: 100 minutes
- Passing Score: 700/1000
- Validity: 1 year (renewal required)
Why This Exam Matters for AI Architects
AI-102 was significantly redesigned in April 2025 to include agentic solutions and generative AI as major topics, reflecting the shift to Azure AI Foundry. It's now the most directly relevant certification for solution architects building AI on Azure.
Skills Measured (April 2025)
| Domain | Weight |
|---|---|
| Plan and manage Azure AI solutions | 20-25% |
| Implement generative AI solutions | 15-20% |
| Implement agentic AI solutions | 5-10% |
| Implement computer vision solutions | 10-15% |
| Implement natural language processing solutions | 15-20% |
| Implement knowledge mining and document intelligence solutions | 15-20% |
How to Study (learning-science approach)
The same evidence-based sequence used across these cert tracks:
- Diagnose first with the official practice assessment to expose weak areas.
- Build in Azure AI Foundry. AI-103 rewards practitioners — implement a RAG app and a multi-agent flow in Python, don't just read.
- Active recall + spacing. After each module, reconstruct the pattern from memory; revisit across days (Dunlosky et al., 2013).
- Teach it back with a short write-up or diagram per domain.
- Re-test to 80%+ on the practice assessment before booking.
Preparation Resources
- AI-103 credential page (free)
- AI-103 study guide (free)
- Azure AI Foundry Documentation (free docs)
Verification & currency
For the full verified certification map (AI-103, AZ-305, SC-500, GH-300, AI-200) and the sources behind these facts, see Fuentes y Verificación v4. AI-102 retired 2026-06-30; confirm AI-103 status and skills measured on the official page before scheduling.