Plan and Manage Azure AI Solutions
- Select the right Foundry services for generative AI and agents
- Set up AI solutions in Foundry: infrastructure, deployment options, connecting to CI/CD
- Operations and security: quotas, scaling, rate limits, cost, managed identity, private networking, keyless operation
- Monitor quality: keep an eye on drift and grounding quality
- Implement responsible AI: safety filters, guardrails, evaluators, trace logging, governance for agents
Implement Generative and Agentic Solutions
- Build generative applications with Foundry: RAG, tool-assisted workflows, evaluation, Foundry SDKs
- Develop agents: roles and tool schemas, function calling, conversation memory
- Multi-agent orchestration, approval workflows and monitoring
- Optimise and operationalise: prompt engineering, self-critique loops, tracing, token and latency analysis, multi-model orchestration
Computer Vision Solutions
- Image and video generation: inpainting, mask-based editing, video editing
- Multimodal understanding workflows: image captioning, visual question answering, alt text, object detection
- Responsible AI for multimodal content: unsafe content, indirect prompt injection via image text, watermarking and brand guidelines
Text Analytics and Speech
- Text analytics with language models: entities, topics, structured JSON, sentiment and tone
- Translation with Azure Translator, adapting it to your own domain
- Speech solutions: speech recognition and speech synthesis for agentic dialogues, Custom Speech, multimodal audio reasoning
Information Extraction
- Retrieval and grounding pipelines: ingestion, indexing, semantic, hybrid and vector search
- Enrichment via skills, OCR-assisted RAG, connecting to agent tools
- Extract content from documents: OCR, layout and field extraction, Content Understanding, structured and markdown analysers
