How I'm Using AI Tools in My Daily Workflow

How AI Tools Are Improving My Daily Workflow

Since the begginging of the AI “bubble” I’ve been reading a lot of people that claim that AI is making them x100 more productive and we programmers are cooked.

While I think that is an exaggeration, it’s true that we are living a new era of productivity for knowledge workers. AI tools have changed the way I approach my daily tasks, helping me research more quickly, code more efficiently, and manage projects more smoothly.

Here’s a look at the tools I use regularly and how they’ve helped me become more productive.

Faster Research with ChatGPT and Brave Search AI

When I’m exploring new topics or facing challenging problems, I rely on ChatGPT and Brave Search AI.

ChatGPT for Reaserch

Whether I’m learning new programming concepts, debugging errors, or planning projects, ChatGPT helps me clarify my thinking through conversation. Its ability to handle follow-up questions makes research feel more natural compared to traditional searches and the new research mode is really good.

Brave Search AI for quick searches

Brave Search’s AI summaries give me quick, concise overviews from across the web. Instead of sorting through multiple websites, I instantly understand key ideas, saving a lot of time. If I need more detailed information, traditional search results are readily available.

Using ChatGPT and Brave Search AI together has significantly reduced the amount of time I spend on initial research.

Supervised Coding with Cursor

Cursor has improved my coding experience by making it easier to write and maintain code. Built specifically with AI assistance, Cursor continues to improve and adapt to my projects.

Helpful Code Suggestions

Cursor understands the context of my codebase, including dependencies and project structure. It provides accurate and practical suggestions, making tasks like refactoring and testing simpler.

Its tab feature is really good and it replaces most of macros for repetitive code editing.

Coding Agent, rules and MCPs

Besides the tab feature, all new generation IDEs have an agent feature, and Cursor is no exception. It’s basically a chat interface that instead of just printing the answer, it executes the action and it will try to apply the changes automatically to the codebase.

I see two important things to consider:

• What tools the agent has access to. For example, Cursor agent can use MCPs to connect to external tools/data.
• What limits and rules we set for the agent. We can write statements in a markdown file that will be used depending on the file type the agent is working with.

MCPs

MCP (Model Context Protocol) is a new open protocol designed to standardize how applications provide context to Large Language Models (LLMs).

It offers a uniform method for connecting AI systems to various tools and data sources.

{
"mcpServers": {
  "postgres": {
    "command": "docker",
    "args": [
      "run",
      "-i",
      "--rm",
      "mcp/postgres",
      "postgresql://user:password@host:5432/database"]
  },
   "mcp-atlassian": {
    "command": "docker",
    "args": ["run", "--rm", "-i", "mcp/atlassian"],
    "env": {
      "CONFLUENCE_URL": "https://your-domain.atlassian.net/wiki",
      "CONFLUENCE_USERNAME": "your.email@domain.com",
      "CONFLUENCE_API_TOKEN": "your_api_token",
      "JIRA_URL": "https://your-domain.atlassian.net",
      "JIRA_USERNAME": "your.email@domain.com",
      "JIRA_API_TOKEN": "your_api_token"
    }
  }
}
}

Database Context

I use the MCP for Postgres to connect to my database and get the context of the tables and relationships. With this tool, I can ask the agent to use the database context to generate the code so the results are far more accurate than if I just ask the agent to generate the code.

Better Collaboration with Jira and Confluence

Integrating Cursor with Jira and Confluence helps streamline my project management and documentation.

I’m trying out the MCP for Atlassian to connect to Jira and Confluence.

Cursor is now able to understand Jira tickets and aligns coding tasks with specific requirements. It also generates well-structured documentation in Confluence, making collaboration easier and more efficient.

This integration helps reduce the mental load when switching between tasks.

Custom Project Rules

I define project guidelines, coding standards, and specific terminology within Cursor rules. This helps Cursor provide suggestions that closely match my coding style and project requirements.

Looking Ahead

I’m excited about future improvements in AI tools, especially:

• Better integration with version control systems. For a fully automated workflow, the agents will be able to read the issue from ticketing system, implement the solution and push the changes to repository.
• AI-supported decisions in software architecture. Better models and tools will make them able to take part in the software architecture decisions. I don’t think we are there yet, but it’s a good direction.

Balancing automation with human oversight remains crucial, at least for now, helping me stay actively involved while using AI to handle repetitive tasks.