Backend-focused engineer building reliable C#/.NET services and AI-102-aligned Azure AI implementations: Vision (Image Analysis 4.0), Video Indexer, and GenAI observability.

I’m a Senior .NET Developer based in Canada, working remotely. My core focus is backend engineering: designing APIs and services with C# / .NET, with strong emphasis on performance, maintainability, observability, and safe delivery.

I’m currently going deeper into Azure and Azure AI, turning each study topic into a small, operable backend project, aligned with the AI-102 certification.

What I value in engineering: clear code, measurable decisions, data over opinions, and systems that are easy to operate in real life.

My workflow is intentionally simple and repeatable:

1. Topic Extraction → Identify weak exam areas and extract study-ready topics
2. Focus → Study one capability deeply (no multitasking)
3. Practice → Answer exam-style questions and implement the capability
4. Consolidation → Document learnings, add edge cases, and convert to reusable notes

This keeps learning practical, measurable, and aligned with real-world backend work.

• C#, .NET, ASP.NET Core
• REST APIs, background processing, integration services
• Entity Framework Core, Dapper (when appropriate)
• Authentication / Authorization (JWT, OAuth concepts)
• Performance tuning, caching patterns, async/await best practices

• Microsoft Azure
• App hosting & compute (App Service / Functions)
• Storage patterns (Blob Storage)
• Azure AI Services (AI-102 track)
  • Azure AI Language (text analysis, custom classification, NER, Q&A, CLU)
  • Azure AI Vision (image analysis, detection, classification)
  • Azure AI Document Intelligence (structured document extraction)
  • Azure AI Search + embeddings (RAG fundamentals, retrieval patterns)
  • Azure AI Content Safety (content filtering and policy decisions)

• Git / GitHub
• CI/CD basics
• Clean Architecture (pragmatic use)
• Testing mindset (unit + integration)
• Observability mindset (logging, diagnostics, monitoring)

Projects listed here are practice snapshots aligned to the AI-102 exam. Focus shifts based on retake priorities.

A REST API for core Azure AI Language text analysis capabilities, designed as an AI-102 practice service with clean contracts and predictable outputs.

Current study focus:

• Analyze text with Azure AI Language

What I’m practicing:

• Language detection
• Sentiment analysis
• Key phrase extraction
• Prebuilt entity recognition

Maintenance note:
This project is paused and kept as a reference snapshot. Enhancements will resume later.

Tech: .NET, ASP.NET Core, Azure AI Language
Status: ⏸️ Paused (not actively maintained)

A .NET REST API wrapper for Azure AI Vision Image Analysis 4.0, focused on correct SDK usage, REST feature flags, defensive parsing, and throttling resilience (429).

What I’m practicing:

• AnalyzeAsync overloads (Uri vs BinaryData) + correct parameter order
• VisualFeatures flags (Caption | Read | Objects) mapping from requirements
• REST features= requests + parsing nullable/empty fields
• 429 throttling: exponential backoff + jitter + concurrency limiting

Tech: .NET, ASP.NET Core, Azure AI Vision (Image Analysis 4.0)
Status: 🛠️ Active

A backend workflow for Azure Video Indexer: ARM access token generation, async upload/index, polling, and insights parsing.

What I plan to practice:

• ARM generateAccessToken (POST) + permissionType + scope (Account/Project/Video)
• Async state machine: token → upload/index → poll → insights
• Time-based outputs: segments/timestamps + keyword filtering
• Defensive parsing for null/empty lists in insights

Tech: .NET, Azure AI Video Indexer (ARM), Azure Storage (optional)
Status: 🧭 Planned

A minimal telemetry/tracing approach for GenAI workloads: correlation, latency, token usage, trace records, and feedback loop.

What I plan to practice:

• Minimum viable telemetry: latency, tokens, correlation/request-id, alerts
• TraceRecord persistence: prompt/params/output/feedback
• Tracing with Azure Monitor / Application Insights (OpenTelemetry)

Tech: .NET, Azure Monitor / Application Insights, Azure OpenAI
Status: 🧭 Planned

Backlog ideas for end-to-end AI-102 systems. Implementation timing is TBD.

A focused lab to practice prompt flow, prompt templates, and evaluation flows/metrics (quality, groundedness, safety checks) using repeatable datasets.

Focus: Prompt Flow + Evaluation (models & flows) + Prompt Templates
Tech: Azure AI Foundry, Prompt Flow, Azure OpenAI, Azure Monitor / Application Insights (optional)

A grounded RAG assistant that answers from curated content while enforcing safety policies.

Focus: RAG + Azure AI Search grounding + Azure OpenAI prompting + Content Safety
Tech: .NET, Azure AI Search, Azure OpenAI, Azure AI Content Safety, Blob Storage

A backend-first agent that selects tools, executes tasks, and returns structured results with observability and safe tool usage.

Focus: Agents (tool calling) + Orchestration + Retrieval tool + Observability
Tech: .NET, Azure OpenAI, Azure AI Search, Functions / App Service

A backend API that accepts text + images and returns Allow / Review / Reject decisions with consistent policy rules.

Focus: Content Safety policy + Vision inputs + Allow/Review/Reject decisions
Tech: .NET, Azure AI Vision, Azure AI Content Safety, Azure OpenAI (optional)

A pipeline that ingests documents and produces structured extraction + semantic insights + summaries.

Focus: Document Intelligence extraction + Vision enrichment + Summarization
Tech: .NET, Azure AI Document Intelligence, Azure AI Vision, Azure AI Language, Azure OpenAI, Blob Storage

• AI-102 retake preparation focused on my weakest areas:
  • Vision: Image Analysis 4.0 (SDK + REST)
  • Video Indexer (ARM tokens + async indexing workflow)
  • GenAI observability (telemetry + tracing + feedback loop)
• Building backend-first implementations: clean contracts, defensive parsing, and resiliency (429)

• AZ-900 (Azure Fundamentals)

• AI-102 (Azure AI Engineer Associate)

• GitHub: https://github.com/adiazjohnson213
• LinkedIn: https://www.linkedin.com/in/arthur-diaz-johnson/
• Location: Canada 🇨🇦 (Remote)