6 Game-changing Automations Every PM Should Try
Automate Communication: Project managers can enhance client communication speed while maintaining a personal touch through automated responses.
Streamline Updates: AI agents can be employed to convert technical project data into clear, client-facing updates.
Maintain Clean Data: Automating data cleanup can prevent issues arising from incorrect or outdated information in CRMs.
Efficient Handoffs: Automated workflows can facilitate seamless transitions between sales and operations, minimizing manual entry.
Judgment Matters: Automation is beneficial for routine tasks, but human oversight is essential for final decision-making.
Project managers are no longer waiting for software vendors to build the workflow they need.These days, many PMs are building it themselves — chaining APIs, scripts, and AI agents together to eliminate busywork, while keeping a human firmly in charge of judgment calls and final approvals.
We asked the operations leaders we knew who were deep in the world of workflow automations: what was the most game-changing one you built into your workflow?
Here’s what they told us works. And where they drew the line on automation.
Unlock for Free
Create a free account to finish this piece and join a community of forward-thinking leaders unlocking tools, playbooks, and insights for thriving in the age of AI.
Have an account? Log In
Automate Client Communication Without Losing the Personal Touch (or the Business)
Many businesses live and die by how fast they respond to a new lead. But it’s unrealistic to expect a human to respond within a few minutes. Until automation steps in, drafting replies automatically, without ever sending anything unreviewed.
All of these examples of automating client communications, show that you don’t need one expensive tool to make automation happen. You can make automations work with your existing stack.
Ready-to-Send Email Drafts
Jacob Varghese, Owner of Zion Foodtrucks, described the pipeline he built to handle inbound leads: “Here is how it is wired: A scheduled job runs in GitHub Actions, so it lives in the cloud and keeps working whether or not my laptop is on. On each run, a Python script pulls new lead emails out of Outlook through the Microsoft Graph API, reads what the person wrote, and hands it to Claude. Claude drafts a reply tailored to the context of their email and their questions, with a Calendly link so they can book a call on the spot. That draft lands back in my Outlook as a real draft for me to read and send.”
Staying on Top of WhatsApp Messages
Walid Chaya, a real estate agency owner, arrived at the same conclusion after a costly delay. “One day a client came in through WhatsApp. I saw the message and didn't feel like replying right away. About two hours went by, and I sent him a message. What that client told me left me shocked: he had already found a property. In two hours! I delayed communication on a whim, and he had already gotten the job done. That's when I realized the idea isn't to reply faster whenever I happen to see it, because, well, I'm human. The idea is to have it automated.”
He explained the simple way he did this: “I took all the social networks I use, which in this case are the ones Meta runs: WhatsApp, Instagram and Facebook. I connected them with an API and an LLM and I linked it all to a calendar.”
One day a client came in through WhatsApp. I saw the message and didn’t feel like replying right away. About two hours went by, and I sent him a message. What that client told me left me shocked: he had already found a property.
Communications Organized By Urgency
Eric Turney, Owner of The Monterey Company, built a system for catching opportunities that show up outside a typical inbox altogether: “F5Bot and Google Alerts send new mentions and relevant discussions by email. Those opportunities are added to a shared Google Sheet with the thread URL, product category, buyer question, urgency, and status, then ChatGPT helps summarize the discussion and prepare a first draft. A person reviews every response for accuracy, tone, and whether replying would actually be useful.”
Turn Scattered Project Data Into Client-Ready Updates
Keeping clients updated on project progress often means pulling status from half a dozen tools and translating it into something non-technical people can actually read.
Jeremy Belcher, Founder of Robot Heart Studios, solved this by building an AI agent that does the pulling and translating for him. “What I automated: the whole release note. I built an AI agent (a Claude skill) that pulls the week's completed work from Linear, checks it against merged GitHub PRs, checks Slack, Email, and my call recordings, then drafts a client-facing release note in plain language and writes it into the client portal. I kick it off, it drafts, I review and post.”
I built an AI agent (a Claude skill) that pulls the week’s completed work from Linear, checks it against merged GitHub PRs, checks Slack, Email, and my call recordings, then drafts a client-facing release note.
He also gave us a rundown of how he connects his toolstack: “Linear's API provides the issue and cycle data, GitHub provides the merge history, Google for email, Slack and Granola MCPs for context, and a Claude agent in the middle translating developer-speak into something a client can actually read.”
The consistency of the output matters as much as the time saved, Belcher noted: “The skill is version-controlled, so it runs the same way every week instead of me reinventing the format each Friday.”
Keep CRM and Task Data Clean and Self-Updating
Dirty data quietly undermines more projects than most teams realize, and a few practitioners have automated the cleanup itself rather than just the reporting on top of it.
Self-Updating Client Data
Justin DeBron, General Manager of Peak Time Pickleball, skipped the drag-and-drop automation tools entirely and went with a small custom script that runs on an automatic schedule: "I deliberately skipped no-code middleware. It is a Python script running as a scheduled GitHub Actions workflow on a cron trigger. Every segment in HubSpot is now self-updating instead of hand-built. The job also surfaces data problems on its own. On the first clean run it flagged 259 contacts with no email and a few typo addresses for my team to fix."
Every segment in HubSpot is now self-updating instead of hand-built. The job also surfaces data problems on its own.
Automated Sales → Operations Handoffs
Hamza Amir, Founder of Jabwewed, automated the handoff between sales and operations so that onboarding tasks no longer depend on someone remembering to create them: “I built a Make.com workflow that triggers automatically when a vendor's status updates in HubSpot. It creates a structured onboarding task in ClickUp with pre-filled fields pulled directly from the HubSpot record — vendor name, category, contact info, priority tier — so nothing has to be manually re-entered.”
Where Automation Should Stop and Judgment Should Take Over
Nearly every practitioner we spoke with drew a firm line between what gets automated and what stays with a person, no matter how capable the automation becomes.
Philip Ruffini, CEO of Hire Overseas, told us his rule is, “automate the prep and the grunt work, and let the AI do the first 80 percent. Keep a human on the last 20% and on the actual final call. We never automate the decision, just everything leading up to it.”
Graham Mann, Founder of SEOTakeoff, framed the same principle around what makes automation actually useful rather than just impressive: “My rule: automate sorting, summarizing, and routing, but keep judgment, prioritization, and approvals with a person. That is usually where automation helps project managers work instead of creating a second system people have to manage.”
For project managers, that's the real opportunity: not building systems that run themselves end to end, but building systems that hand you the decisions that matter.
Want more insights like these? Sign up for a free DPM account to hear from more experts like these.
