More How I AI episodes

How I turned Claude into a self-improving PM assistant | Daniel Blum (PM, Melio) thumbnail

How I turned Claude into a self-improving PM assistant | Daniel Blum (PM, Melio)

Published 31 Aug 2026

Duration: 00:46:07

"AI tools dramatically boost productivity by automating tasks, enhancing decision-making, and managing communications, while AI-powered systems scale workflows, adapt to work styles, and reduce cognitive load, with future potential for handling coordination tasks autonomously."

Episode Description

Daniel Blum is a product manager at Melio, a B2B payments company, and one of the most systematic thinkers I've had on the show when it comes to perso...

Overview

The podcast discusses how AI tools can significantly enhance personal productivity, particularly for roles like product managers, by streamlining workflows and reducing time spent on routine tasks. The speaker describes using AI systems - such as Claude and Cowork - to manage 70 - 80% of computer-based work, including processing emails, Slack messages, meeting transcripts, and calendar inputs. These systems autonomously maintain a Notion board, organize weekly priorities, identify action items, and prepare daily briefings, allowing tasks that previously took a week to be completed in a single day. A key aspect is the AI's ability to ask clarifying questions, retain context, and adapt to individual work styles, creating a self-improving, self-optimizing workflow.

The discussion also covers the scalability of these AI systems beyond individual use to benefit entire teams, exemplified by the development of "Workstation," an organizational operating system that simplifies onboarding and standardizes AI integration across roles. The speaker emphasizes the importance of building feedback loops - where AI learns from revisions, user interactions, and recurring tasks - to continuously refine outputs and develop reusable skills. Challenges such as information overload and AI-induced frustration ("Claude rage") are acknowledged, with strategies like critical auditing of new tools and improving system trust over time. Ultimately, the focus is on leveraging AI not just for speed, but for deeper, more strategic work by reducing cognitive load and automating coordination-heavy processes.

What If

  • What if you automated your weekly planning using AI to synthesize inputs from all your tools?

    • Move: Set up an AI agent (e.g., Claude via Cowork) to automatically pull data from your calendar, email, Slack, and Notion every Sunday, then generate and update your "This Week" and "Top of Mind" sections.
    • Why Now?: Manual weekly planning is repetitive and time-consuming; AI can now contextualize cross-platform inputs reliably, and delaying adoption prolongs inefficiency.
    • Expected Upside: Reclaim 3 - 5 hours per week, ensure no action items are missed, and start each week with a coherent, AI-validated plan.
  • What if you trained your AI to flag and clarify ambiguous terms in real time during communication?

    • Move: Build a personal context file with internal jargon, goals, and key projects, then configure your AI (e.g., Claude) to interrupt and ask for clarification when it encounters undefined terms in Slack messages or drafts.
    • Why Now?: Miscommunication due to undefined terms slows down execution, and AI now has the memory and prompting capability to act as a real-time consistency checker.
    • Expected Upside: Reduce rework from misunderstandings by 30 - 50%, accelerate onboarding of new collaborators, and strengthen documentation accuracy.
  • What if you delegated your daily coordination tasks (e.g., status updates, meeting prep) to an AI co-pilot?

    • Move: Implement a daily AI briefing automation that reviews your inbox, Slack threads, and upcoming meetings, then generates a one-page prep sheet with action items, talking points, and context summaries each morning.
    • Why Now?: Coordination overhead consumes 50 - 70% of a solo operator's day, and AI tools like Cowork or Claude Enterprise now support reliable, agentic summarization and prioritization.
    • Expected Upside: Free up 2 - 3 hours daily for deep work, reduce mental fatigue from task-switching, and maintain consistent follow-through across projects.

Takeaway

  • Set up a personal AI co-working system using tools like Claude and Cowork to manage 70 - 80% of computer-based tasks, integrating with daily tools (Slack, email, calendar) for automated task tracking and context awareness.
  • Build a dynamic Notion board structured into "Top of Mind," "This Week," and "Inbox" sections, and automate its maintenance using AI to sync inputs from communication platforms and meeting transcripts.
  • Implement a weekly prep automation that aggregates data from multiple sources (calendar, Slack, Notion) to generate a prioritized task list and auto-update your workflow system at the start of each week.
  • Develop a feedback loop where AI reviews past outputs (e.g., drafts, decisions) against final versions to identify patterns, suggest reusable skills, and improve future performance over time.
  • Use AI to audit new tools and productivity advice (e.g., via a custom "Improve" skill) to filter out hype, assess practicality, and decide whether to adopt, adapt, or discard external recommendations.

Recent Episodes of How I AI

5 Aug 2026 Build an AI code review bot in 30 minutes with Vercel Eve

"AI-powered automated PR review systems streamline software development by scoring and auto-approving low-risk changes, boosting throughput, reducing approval times, and maintaining compliance, as demonstrated by tools like Vercel's Eve and industry adopters like Intercom."

22 Jul 2026 Computer & browser use in Codex (5 real examples)

"AI boosts productivity by automating browser and computer tasks, reducing manual effort through commands like `@browser` or `@computer`, with applications in software testing, personal tasks, and workflow optimization, though human oversight and speed limitations may apply."

More How I AI episodes