AI-Powered Project Manager and Educator Shares a Simple AI Workflow That Fundamentally Changes How Projects Operate

Mashhood Ahmed is an AI-powered project manager, speaker, and educator. In his work, he consistently identifies game-changing opportunities for AI-assisted delivery.

Read Now

By Faye Wai • Last updated on Dec 23, 2025

Key Takeaways

Turning meetings into structured data: Mashhood’s biggest breakthrough wasn’t a complex system — it was meeting transcription + AI prompting. Converting every discussion into searchable, structured data powers faster summaries, clearer action items, earlier risk detection, and more aligned project documentation.

AI now acts as a “silent team member": With tools like Copilot, ChatGPT, and agentic workflows, Mashhood has shifted from manual documentation to a lightweight, AI-assisted delivery model. AI drafts scopes, charters, RAID logs, and even flags risks across multiple meeting transcripts.

AI accelerates delivery: AI can stitch together insights across meetings, highlight contradictions, and draft documents, but it cannot interpret politics, cultural nuances, or stakeholder sensitivities. Mashhood emphasizes a “human in the loop” guardrail to prevent over-reliance.

In our interview, Mashhood shared one workflow in particular that is changing the game, and it's simple. Here are the details.

How AI is redefining the role of the modern project manager

I’m an AI-powered project management consultant and trainer. I’ve delivered 60+ speaking engagements and AI masterclasses in the last two years. I also help teams integrate Gen AI into everyday project workflows to improve speed, clarity, and decisions.

I originally expected AI to help with small efficiencies like formatting or summarizing, but the real impact came from how it connects dots across meetings, emails, vendor documents, and more. Once I started leveraging transcriptions with Copilot prompts, I realized AI can surface patterns, contradictions, risks, and decisions that even the project team misunderstood or didn’t remember discussing.

Because of this, my role as a project and program manager has shifted toward orchestrating AI-assisted decision-making, planning, and delivery. Instead of manually stitching together data from meetings, documents, and systems, I now use AI tools like Copilot and ChatGPT to automate information flow and generate real-time insights.

I deal with lots of meetings and follow-ups as a project leader. It's my job to improve accuracy and alignment in project delivery — and at the same time reduce the number of meetings. That requires an AI-assisted delivery model.

Instead of manually stitching together data from meetings, documents, and systems, I now use AI tools like Copilot and ChatGPT to automate information flow and generate real-time insights.

Mashhood Ahmed

Founding Director at PMassistant.ai

How turning meetings into structured data unlocks faster, clearer project delivery

The most underrated and impactful features of AI in project management have been meeting transcriptions and AI assistants like Microsoft Copilot — working together. It sounds simple, but it fundamentally changes how projects operate.

By turning every meeting into structured, searchable data, it becomes the backbone for almost every downstream project activity. Once the transcript exists, I can do magic on the fly.

In terms of impact, it:

Saves roughly two hours per day in documentation and follow-up — time and energy I now spend on strategic planning, stakeholder alignment, and AI and project governance.
Improves project alignment because everyone has access to the same source of truth.
Surfaces risks and scope gaps earlier before they affect scope, time, or budget.
Reduces unnecessary meetings because documentation is clearer and complete.
Strengthens transparency and accountability across the team.

For example, during a healthcare implementation, a series of five requirement meetings across different clinics felt disconnected. But with a single prompt, AI stitched all five transcripts into:

A unified requirement set
A clear in-scope and out-of-scope list
Workflow differences between clinics
Early warnings about conflicting processes

This level of insight would have taken days manually, and honestly, some of it would have been missed entirely. I'll dive deeper into that example in a moment.

How an AI-assisted delivery model accelerates delivery

This has led to me shifting from traditional, document-heavy project management toward a lightweight, AI in project delivery model. It doesn’t replace core project management principles; it streamlines how we apply them.

Old approach:

Manually writing meeting minutes
Maintaining separate Word/Excel documents
Chasing people for clarifications
Rewriting status reports from scratch

New lightweight system:

Every meeting is transcribed using Microsoft Teams or Google Meet.
AI productivity tools(Copilot, Gemini, or ChatGPT) generates summaries, action items, risks, and decisions.
These outputs are stored automatically in SharePoint or Drive.
Documents like charters, RACI, communications, and RAID logs are generated from the same transcript data.

Result: Project initiation meetings now generate half the project documentation automatically, reducing manual effort by 40–60 percent.

How AI is reshaping project rituals and reducing manual work

As far as rituals, this use of AI has completely reshaped them. We've shifted manual, meeting-driven processes to continuous, data-driven workflows.

Instead of relying on people’s notes or memory, we now treat AI as a silent team member that listens, synthesizes, and highlights what matters. Here is an example for drafting project scope and charter documents, which are foundational to any project.

Old approach: Scope was drafted manually after a series of meetings, often based on fragmented notes.

New AI-enabled approach:

I feed multiple meeting transcripts, emails, and requirement discussions into Copilot or ChatGPT.
AI drafts the first version of the scope, including in-scope, out-of-scope, assumptions, and dependencies.
I refine it with SMEs instead of building it from scratch
I leverage AI thinking to draft project scope and charter within a few hours instead of days

Mashhood's tip

Instead of relying on people’s notes or memory, we now treat AI as a silent team member that listens, synthesizes, and highlights what matters.

How agentic AI is automating project workflows and reducing repetition

Over the last six months I’ve also been actively experimenting heavily with agentic workflows and orchestration-style platforms to reduce repetitive PM tasks and create more consistent delivery patterns. Microsoft Agent 365, released in mid-November, has especially strong potential for multi-agent orchestration in real project environments.

One example is a custom-built Meeting Agent that I use for the meeting workflows that I mentioned above. The agent automatically joins calls, generates meeting summaries, and produces structured action item tables with owners and deadlines. Those action items can then be pushed directly into Jira, Asana, or other work management tools, eliminating manual transcription and follow-up.

We also set up a Risk Agent that analyzes meeting transcripts to flag new risks immediately after discussions. It allows the project manager to simulate scenarios, evaluate potential impacts, and decide whether the risk needs escalation or team visibility.

These small but powerful agentic workflows are already reducing administrative workload and helping teams operate with more consistency, transparency, and early-risk awareness.

A real-world example of AI-driven alignment across complex projects

Let's dive deeper into that example of the multi-location healthcare clinics. The project had tight timelines, multiple vendors, and frustrated stakeholders due to inconsistent requirements.

The challenge

We had five locations, each with different workflows, and lots of meetings with clinicians and stakeholders at different levels. Requirements were not consistent from one location to another, and the team was losing alignment because decisions for each location were buried — mainly in meeting notes and sometimes in emails and chat messages.

The AI-driven approach

As a first step, we set up centralized meeting intelligence. We turned on Microsoft Teams transcription for every meeting and used it as a single source of truth.

We then built simple AI in team collaboration automations to store transcripts on a shared drive or SharePoint. And we provided access to the whole team, so that anyone could interact and prompt multiple meeting transcripts at a time.

For example, I often prompted:

"Based on this meeting transcription, help me generate a meeting summary under 200 words and an action item table, so that as a project manager I can follow up on the agreed action items, resources, and deadlines."
"Based on this meeting, help me identify any new risk or help me draft a risk mitigation plan as proposed by the technical architect."

By enabling meeting transcription and simple prompts like this, we eliminated the need for manual note-taking and the review of scrambled, handwritten notes. We also stabilized requirements from different clinics and their required workflows.

For example, I can identify requirement gathering meetings and then simply prompt, "Help me write high-level requirements to document and finalize project scope, make sure to include in- and out-of-scope separately, as discussed in these meetings."

Boom. Here I have a project scope draft ready based on these five meetings from each clinic. Further, a follow-up prompt like, "Highlight workflow differences between the five clinics and suggest how we can streamline them" would help identify differences that we can discuss with clinicians to streamline and follow up.

The benefits

It's worth noting that, initially, stakeholders had reservations. But once we did the pilot and showed the value, almost everyone bought in. Further, team members became careful about what they said in those meetings, which was eventually good for the project and the business. Here's the feedback that stood out:

AI in stakeholder management helped identify edge cases and workflow variations that usually never get documented or discussed in early meetings. These became visible immediately because we were prompting across transcripts instead of relying on memory or handwritten notes. In a few cases, this even expanded or clarified project scope.
Stakeholders appreciated having a single source of truth. Instead of digging through emails, chat threads, or personal notes, they could go to one place and see decisions, follow-ups, and actions clearly organized.
Business and technical teams noted that AI surfaced risks and process gaps earlier than our traditional methods. Items like missing workflows, inconsistent roles, unclear responsibilities, or potential delays were flagged right away based on meeting data.
Teams appreciated that AI-generated summaries and action items reduced prep time. They walked into meetings with more clarity and fewer misunderstandings.

Without human judgment, AI is fast… but not necessarily right.

How to maintain a high-quality knowledge base with AI transcription

These days, transcripts are highly accurate, but challenges still occur with strong accents, fast speakers, overlapping conversations, or unstable internet. Here’s how I manage accuracy in practice:

Set expectations early. We let the team know that transcriptions are a support tool, not a verbatim legal record. This removes pressure and encourages open participation.
Build a "clarification loop" into our process. After each meeting, team members are encouraged to review the summary or action item table and flag corrections. If any statement looks off, ambiguous, or contradictory, we validate it in follow-ups or the next meeting. This takes just a minute or two and greatly improves accuracy.
Prompt AI correctly. We always prompt AI tools to interpret context, not to assume everything is literal. For example: "Based on the transcription, summarize the key points. If anything is unclear, highlight it rather than assume."
Encourage better meeting habits. Clear turn-taking, avoiding cross-talk, and slowing down a bit when discussing decisions all help improve transcription quality and improve meeting quality overall.

Why Microsoft and Google’s AI stacks are becoming core to project delivery

Here are the two technology stacks I've used with clients:

Microsoft stack: We use Teams, Outlook, Copilot (enterprise), SharePoint, OneDrive, and MS Office. We even use some Power Apps and Power Automate to build integration with AI project management tools like Smartsheet or Asana.
Google stack: Google Meet, Gmail, Calendar, Gemini, Drive, Google Docs and Sheets. In this case, we use Zapier for PM tool integration and workflow automation software. In the last few months, the Google stack has become much more AI-capable, especially with the rollout of Gemini for Workspace and Vertex AI integrations.

Since Microsoft and Google compete in the corporate space, their AI capabilities have greatly evolved in the last year. There are a number of other productivity tools we could be using; however, I am betting on Google and Microsoft for now.

Why over-reliance on AI is a new project risk

I was surprised by the cultural shift I saw when implementing AI in project analytics. Once the team experienced how reliable AI outputs could be, people became more disciplined about articulating decisions clearly in meetings because they knew the transcript would be used as a source of truth. It improved project discipline without me having to enforce it.

Also, team members who had initial reservations started leveraging Copilot for everyday tasks. In fact, once teams see AI’s value, over-reliance becomes the new risk. I now mitigate this risk with human guardrails — i.e. a human in the hoop to review.

How to decide what project work to automate with AI — and what requires human judgment

Gen AI definitely helps accelerate work. However, human judgment is still required — you can't say to the steering committee that "Project scope was written by Gen AI." It has your name, thus you must take ownership and apply human judgment.

Here are the aspects of delivery work that I think are most ripe for AI:

Meeting intelligence
Drafting project documentation
Identifying project risks earlier
Document analysis, like SoW and SLA from vendors and how these documents are aligned or not aligned

And recently, we had some success in drafting documents using the client's document template, so the response generated by Gen AI has a proper document structure.

AI in project decision making does not understand politics, cultural nuances, or stakeholder sensitivities. That’s where human expertise becomes essential. We must evaluate:

Which stakeholder’s opinion carries more weight
Where resistance may occur
How decisions affect other programs or departments
What compromises are acceptable

Without human judgment, AI is fast... but not necessarily right.

mashhood's tip

AI is well on its way to becoming the de facto project management layer, fundamentally reshaping how projects are planned, monitored, and delivered in the near future.

Why AI is becoming the new project management layer for modern organizations

AI in project planning is here to stay. With tools like Copilot and Gemini now deeply integrated into email, calendars, documents, and core business applications, AI is well on its way to becoming the de facto project management layer, fundamentally reshaping how projects are planned, monitored, and delivered in the near future.

Agentic AI has enormous potential too, and while it's too early to predict its full impact, it could become a significant disruptor for many traditional businesses.

What’s clear is that organizations that adapt early will gain a major advantage in speed, efficiency, and decision-making.

Why you should start now with a small-scale AI pilot

My advice is simple: don’t wait for perfect clarity, start experimenting now with a small-scale pilot.

Here are a few practical recommendations:

Start small, but start now. Turn on meeting transcription. Ask AI to summarize a status update. Use Gen AI (Copilot or ChatGPT) to collaboratively draft a project document. Small experiments build confidence and momentum.
Treat AI as a teammate, not a threat. AI won’t replace delivery leaders, but delivery leaders who use AI will outperform those who don’t. Use AI to handle the heavy lifting so you can focus on alignment, coaching, negotiation, and strategy.
Invest in human judgment and governance. AI can surface insights, but humans still make the decisions. Strengthen your skills in risk management, stakeholder engagement, ethical considerations, and project governance. These skills will matter more than ever.