Portfolio Assistant: Project Management Summary

Project Overview

Description: A conversational assistant developed to answer questions about my portfolio and qualifications using cloud computing and AI services.
Code: GitHub Repository
Approach: Given the planned timeline (4-5 days), I am using an iterative, Agile-inspired approach. I am focusing on immediately gathering requirements, planning the core features, and establishing what cycles of implementation and review may be required. At the end of each cycle, I will assess progress, and leave room to refine the plan as needed.
End Result: Web Portfolio Chat Interface
Phases:
- Requirements gathering
- Design and architecture
- Implementation
- Testing and QA
- Deployment
- Review and retrospective

Project Log

2025-09-11

Phase: DefinePhase: MeasureDay: 1Status: donescopingrequirementslearning

Day 1: Requirements Gathering, Learning, and Planning

Goals:
- Define project requirements and constraints
- Identify key features/functionalities
- Research relevant technologies
- Draft initial plan
Outcomes:
- Set the hard limit for project completion to 5 days
- Defined key features:
  - Conversational interface (ultimately for use on webpage)
  - Responses informed by documents that I provide
  - Deployment via GitHub Actions (including linter and tests)
- Identified key constraints:
  - Limited time (4-5 days)
  - Limited budget
- Selected Azure Functions due to good documentation and fairly comprehensive free trial of Azure services
- Started basic course on Microsoft Azure Fundamentals
- Created a storage account
- Drafted initial project plan:
Next Steps:
- Finalize system architecture
- Set up Azure resources
- Define a preliminary file structure

2025-09-12

Phase: MeasurePhase: AnalyzePhase: DesignDay: 2Status: donelearningarchitecturedevelopment

Day 2: Finalize System Architecture and Set Up LLM Resource

Goals:
- Set up conversational AI resource
- Finalize system architecture
- Begin development of chat agent
- Set up GitHub repository
Outcomes:
- Completed Azure Fundamentals Course
- Set up Azure AI Foundry resource (GPT-4.1)
- System architecture:
  - Azure Function App to coordinate resource interactions
  - Azure AI Foundry for language model
  - Azure AI Search for context retrieval
  - Azure Blob Storage for document storage
  - GitHub Actions for CI/CD
- Set up local development environment (Python, Azure Functions Core Tools)
- Set up basic file structure
- Began development of (agent_client.py) AgentClient class
- Began development of (main.py) function app
Backlog:
- Set up GitHub repository
Next Steps:
- Set up function app
- Continue development of chat agent
- Test chat functionality locally
- Deploy function app

2025-09-13

Phase: DesignPhase: VerifyDay: 3Status: donetestingdevelopmentdeployment

Day 3: Deploy Function for Agent and Create Basic Chat Interface

Goals:
- Create Azure Function App
- Develop chat agent
- Begin local testing
- Set up GitHub repository
- Deploy function app
Outcomes:
- Created Azure Function App resource
- Performed initial unit tests for AgentClient and main.py
- Performed initial static code analysis with flake8
- Restructured project files for compatibility with Azure Function App (src/portfolio_assistant/main.py -> chat_function__init__.py)
- Started local testing of chat agent functionality
- Resolved local issues (authentication, query formatting)
- Created deploy.yml for GitHub Actions CI/CD
- Created GitHub repository and pushed initial code
- Resolved deployment issues (environment variables, package contents)
- Added GitHub secrets for function app
- Set up basic user interface to test deployment
Next Steps:
- Continue refinement of chat agent
- Continue testing and debugging
- Start on AI Search implementation/integration

2025-09-14

Phase: ImproveDay: 4Status: donesemantic searchtestingintegrationdevelopment

Day 4: Integrate Semantic Search into Chat Agent

Goals:
- Set up Azure AI Search resource
- Add reference document base to Blob Storage Container
- Integrate semantic search into chat agent
- Deploy integrated search functionality
Outcomes:
- Created Azure AI Search resource
- Added documents to Blob Storage Container
- Developed SearchClient class
- Updated chat function to incorporate SearchClient
- Performed unit tests for SearchClient and chat function
- Tested integrated search functionality locally
- Resolved integration issues (authentication)
Backlog:
- Deployment of chat agent with integrated search functionality
Next Steps:
- Continue development of chat agent with integrated search
- Conduct further testing and QA

2025-09-15

Phase: ImprovePhase: ControlDay: 5Status: donetestingUATQAdevelopmentdeployment

Day 5: Deploy Agent with Integrated Search

Goals:
- Deploy chat agent with integrated search functionality
- Conduct comprehensive testing and QA
- Refine agent performance
- Prepare for going live
- Estimate cost of continued support beyond free trial
Outcomes:
- Deployed chat agent with integrated search functionality to Azure Function App
- Addressed code linting issues
- Performed user acceptance testing (UAT)
- Identified issues with response quality
- Iterative Improvements based on UAT
- Deployment of improved chat agent
Backlog:
- Prepare for going live (rate limiting)
- Estimate cost of continued support beyond free trial
Next Steps:
- Plan for future enhancements and maintenance

Retrospective

Throughout the project, following the overall plan but keeping daily tasks flexible was crucial for adapting to any challenges that arose. Extensive local testing and utilizing GitHub Actions for CI/CD significantly streamlined deployment and improved code quality. I will continue to refine the agent and plan to make it available on my portfolio website once proper security measures are in place. Although development on another project was disrupted, it was a worthy trade-off for the opportunity to learn about cloud computing and AI services, which has opened many new possibilities for augmenting my work.

Key Project Management Skills Demonstrated

Rapid requirements gathering/scope definition
Iterative planning and Agile-inspired execution
Time and resource management under tight constraints
Risk identification and mitigation (e.g., handling deployment issues)
Daily review and backlog management
Clear documentation of goals, outcomes, and next steps

Project Log, Continued: Revisiting Efficiency

2025-11-09

Phase: ImproveDay: 6Status: donecost optimizationdevelopment

Day 6: Replace AI Search

While it meets the needs of the project, the Azure AI Search is not really being utilized to its full potential. It would be smart to look into more cost-effective alternatives that still apply advanced search techniques. This is not high priority, as I have the ability to maintain the current infrastructure while I develop an alternative solution. After successful testing, I will switch to the new search client. I will target completion by the end of the month.

Goals:
- Improve cost of integrated search functionality
Outcomes:
- Researched alternative search solutions
- Decided to implement Apache Solr
- Created Azure Virtual Machine to run Solr search

2025-11-10

Phase: ImproveDay: 7Status: donedevelopmentdeployment

Day 7: Integrate Solr Search into Chat Agent

Goals:
- Integrate Solr search into chat agent
- Deploy integrated search functionality
Outcomes:
- Added new indexer script (index_for_solr.py)
- Added new class to search_client.py (SolrSearchClient)
- Updated chat function to use new search class
Backlog:
- Deploy integrated search functionality
Note: I have decided to take on a project with higher priority. I hope to resume integration of the new search client sometime in Spring of 2026.