We ❤️ Open Source
A community education resource
How I use GenAI to prioritize the deluge of email
A smarter way to triage email with context-aware prompts.
This article is part of the eBook: Everyday AI guide: Practical genAI life hacks from real users, a free download from We Love Open Source.
Every morning, I face the same digital avalanche: Dozens of emails competing for attention. Sound familiar? After 25+ years of running CrossComm, I’ve learned that the biggest productivity killer isn’t the volume of communication—it’s the mental overhead of constantly making micro-decisions about what deserves attention first.
This is exactly where large language models (LLMs) excel, and it’s a perfect example of leveraging AI for what it does best: Making lightweight judgment calls that speed up daily workflows.
The power of context-aware prioritization
Traditional email clients offer basic sorting—by date, sender, or subject line. But what if your AI assistant could evaluate emails the same way you would, considering relationship history, business relevance, and contextual urgency? That’s now possible through Model Context Protocol (MCP), accessible to any AI user through ChatGPT’s and Claude’s “Connectors” functionality.
Setting this up requires no technical background. For ChatGPT (paid version required), simply enable Gmail or Outlook connectors in Settings. Claude offers a Gmail connector out of the box.
Important privacy note: Enabling this connector will allow your AI tool to read your email; do not proceed if you are not OK with that. If using ChatGPT, take care to disable content usage for model improvements in your settings; if using the free version of Claude, turn off the “Help Improve Claude” option under Settings > Privacy).
Read more: Deep dive into the Model Context Protocol
My email triage system
Here’s the prompt I use to wrangle my email, feel free to adapt it to your priorities:
Look at the most recent 10 unreplied messages in my email inbox; for each unreplied email, evaluate the importance of replying to that email through the following criteria:
- The importance of the sender and my pre-existing relationship with that sender (senders who I have not send email to before are less important, senders who I have a pre-existing working relationship with or represent a client are very important)
- The importance of the request to my business or personal life
- The urgency of the request
- Deprioritize solicitations for products, services, or investment opportunities
- Prioritize emails regarding prospective/new project work and engagements for CrossComm
- Ignore email notices that are not expecting a response such as automated notices from github or sentry or purchase receipts
- Ignore non-personalized emails that are not directly addressed to me
After you’ve evaluated the importance of these unreplied messages, generate a prioritized task list (most important first) for replying to those emails.
The AI doesn’t just sort, it reasons through relationships, business context, and communication patterns I’ve established over my entire career, if offered the context and opportunity to do so.
Getting unexpected results? Invest in refining your prompt until it judges emails as you would. The above prompt has already gone through many iterations and I’m sure the above will not be my last iteration!
Beyond prioritization
Once you have your prioritized list, push further. Ask ChatGPT or Claude to draft responses, incorporating context from other connected systems like your CRM or project management tools. Pro-tip: The same approach works brilliantly for Slack and Microsoft Teams message management through their respective connectors. Don’t stop with email—wrangle your deluge of chat threads as well!
The real AI advantage
This application highlights where LLMs truly shine. We’re not asking AI to replicate deterministic software capabilities that existed before—basic sorting and filtering have been around for decades. Instead, we’re leveraging AI’s ability to make nuanced judgment calls that were previously impossible to automate.
Email prioritization perfectly demonstrates this strength: Understanding context, weighing relationships, and making decisions that unblock workflows in ways that traditional software simply cannot. That’s the future of practical AI—reducing cognitive load so that our mental attention can be focused on where it matters most.
Get the “Everyday AI” eBook
This free guide highlights simple, real-world examples of how people are using AI to save time, spark creativity, and make life easier.
The opinions expressed on this website are those of each author, not of the author's employer or All Things Open/We Love Open Source.