MyAIAssistant¶

Project Goals¶

MyAIAssistant helps users organize tasks, projects, organizations, persons, and use subject-matter knowledge to support task execution. The application leverages AI for semantic search, note summarization, task extraction and recommendations. The tool links knowledge artifacts to tasks to provide better context when addressing work by using AI help. It is also including an agent management and definition, so tasks can be automated, like getting news of the days, get Slack channel summary...

For task management, the approach is based on Stephen Covey's "7 Habits of Highly Effective People" the system helps manage priorities efficiently using the Eisenhower Matrix (Urgent/Important classification).

Project Principles¶

1. Run locally - All core features work without external LLM API dependencies. Still it is possible to use remote services, but the approach is to define prompt, tools, specific context and run local models.
2. Isolation: Knowledge, project database, vector store can be isolated in a workspace. Application is started within a workspace. One command to create a workspace. Global settings can be done to get some knowledge, skills, ... for cross workspace. Data stays on local infrastructure
3. Agentic solution: Agents, tools, teams, workflows are declared with config file, but can be extended by custom classes. Predefined agents are used to do research, summarization, task recommendations.
4. Efficient prioritization - Eisenhower Matrix helps focus on high-impact work, and tasj integrated to project and organization user may work with. This tool applies well to solution engineers, or consultants who have to manage lot of small engagements. With a drag-and-drop user interface it is easy to continuously re-prioritize tasks. With AI assistant, the tool helps to better run a working day, query what was done last wee, last month, last quarter...

Access the webApp local once started.

Why I built this?¶

There are ton of to do apps, why a new one? The main drivers for creating this app:

• How I can augment myself and my day-to-day outcomes?
• Address what are the things I can do more, related ot what I value? I think the answer is learning, but how to organize time for it, and how to really track, learning activities?
• I used paper or digital notes based on the Eisenhower matrix to track my tasks, but I think, AI can enhance the plannification, the decomposition and the semantic search of my current knowledge, my past work, and projects.
• Address how to maximize my time on desired activities and minimize time on overhead?
• While tracking task with app, it is easy to build weekly report, an essential tool for management.
• NotebookLM, Glean, RAG, GraphRAG... are excellent tools to start managing knowledge and query it, but they are not related to my work directly. I think it may be possible to fill this gap, by linking all the information about organizations, projects, tasks, assets all together.
• The tool should also helps to address the efficient use of AI, by helping the step of clear thinking (getting a clear statement of the problem), clear writing (by formalizing in simple sentences the problem, to explain what you want to do), to build prompts for more efficient AI.
• Apply agentic architecture for a simple day-to-day application, as a learning experience.

Core Features¶

Quick Start¶

• Getting started with workspace and the application.
• Task management user guide.
• Project management user guide
• Organization management
• Knowledge management for better AI context

Understanding the implementation¶

See design and implementation chapter.