AI Workflow: OKR Alignment Tracker

TLDR: I built a lightweight workflow that replaces manual reporting with an AI-assisted system that highlights work aligned to company OKRs. It surfaces overlooked contributions, flags misalignment by department, and gives teams a real-time view of progress without needing status meetings or manual curation.

Tools Used: Notion, Gumloop, Open AI GPT 4-Mini, Slack

—————————————————-

The Problem With Traditional OKR Reporting

Most companies have some kind of structure for weekly updates such as shared Google Docs, Slack threads and standing meetings. These systems are meant to create visibility and keep everyone aligned. But even when they’re consistent, they still rely on what a few people (usually  heads of departments) decide is worth surfacing. What most organizations end up with is a version of progress that may be polished, but isn’t always complete or connected to actual goals.

Most reporting processes are still bottlenecked through people

If you’re a manager, you know the drill. Sunday evening rolls around and instead of relaxing, you’re piecing together updates, scrolling through Slack, scanning docs, trying to remember who said what. You debate whether to include a half-done project, wonder what you’re overselling or underselling, and stress over how to summarize it all in 150 words. When company progress updates depend heavily on who had the time and headspace to write one, that’s a bottleneck disguised as a process.

Cross-team patterns stay hidden

Team reports tend to look inward when summarizing their own progress, blockers, and priorities. Without a shared view across departments, broader patterns stay invisible. Multiple teams might be solving the same problem without knowing it, or one team might be blocked by something another team already resolved. Even alignment to the same OKR can look completely different depending on how each group frames their work. 

OKR reflections need a company-wide lens that reveals macro trends, duplicate efforts, and misalignments individual teams wouldn’t catch on their own.

Traditional reporting also misses early signals of meaningful work

Sometimes, high-leverage work doesn’t show up in the metrics right away. It might be foundational, exploratory, or simply take longer to show results. In traditional reporting, this kind of work often gets overlooked because it’s harder to quantify and easy to dismiss when you’re focused on short-term outcomes. But early signals matter. They show where momentum is building, which bets are starting to pay off, and what’s worth doubling down on. Without a system to catch those signals early, teams risk under-investing in what actually drives progress, and overvaluing whatever happens to be immediately measurable.

Ultimately, what any organization needs today is a new baseline for measuring progress. Instead of people spending hours compiling and curating updates, AI becomes the neutral aggregator that surfaces what’s truly important, allowing humans to focus on making sense of it.

As humans, we instinctively surface whatever we remember, whatever looks finished, or whatever feels politically safe. But you can design AI to not give a fuck. To pull ground truth directly from where work happens instead of serving up carefully crafted narratives. To show you the misalignments and duplicate efforts that no one wants to admit exist.

This frees leaders to do the work that actually matters: removing blockers, reallocating resources, and doubling down on what's working.

In practice, OKR alignment becomes a living system rather than reporting theater. And when an AI layer surfaces information that hurts your feelings, you can skip the defensiveness and channel that energy into actually fixing problems.

How to Set up The OKR Alignment Tracker

To get started, you need:

• Slack workspace

• Notion workspace with:

  • A Notion page with your OKRs (your source of truth for company/team goals)

  • A Notion database where raw submissions go 

  • Another Notion database where the live report is saved

• Claude or Open AI API access for analysis

• Gumloop for workflow automation

Your Gumloop setup will be two separate flows: the reminder flow and reporting flow.

Flow 1: Weekly Reminder

The automation starts with a Slack reminder. On Friday, a short message pings the team with a link to a Notion form. People only need to answer a few questions: What did you work on? What did you learn? What would you like to brag about? 

Because there’s no expectation of polish, it takes much less time to fill out. People can focus on describing work done in their own words and that freedom lowers the barrier to sharing. It also helps capture work that might otherwise go undocumented especially from people who aren’t naturally inclined to self-promote.

This flow is important because the consistent Slack reminder builds a weekly habit and encourages everyone to contribute, not just managers. Even though it links out to Notion, the flow stays lightweight and familiar. And because updates come directly from the people doing the work, the system captures a wider range of activity with less overhead.

Flow 2: Weekly Reporting

The second part of the workflow runs every Monday morning. Behind the scenes, the automation pulls all the submissions from the Notion database with raw, unfiltered entries written by employees the previous week. It submits all compiled entries to your LLM which then compares each entry against the company’s actual OKRs.

This flow generates two outputs: 

The first output is a detailed Notion Report (800-1200 words), which is then saved to a "Weekly Updates" Notion database with:

• Executive summary of the week

• OKR breakdown showing which teams moved each key result forward

• Team recognition section highlighting specific contributions

• Key learnings and insights

• Forward-looking concerns or blockers

• Recommendations to keep the team on track

The second output is a Slack Summary (250-300 words) posted to a specific channel answering:

• How are we performing against OKRs?

• Who's contributing and how?

• What gaps or risks exist?

• Who deserves recognition this week?

• What can we improve?

Take this workflow as a v.1. It requires maintaining three separate Notion assets which adds overhead even if the process itself is automated. More fundamentally, it's still a garbage-in, garbage-out system. While I've factored for typos and poor articulation, someone who submits "worked on stuff" won't magically get proper representation in the progress updates. The quality of output still depends on individuals actually documenting their work. There's also friction in the submission process where people context-switch from Slack to fill out a Notion form, and this creates just enough resistance that some won't bother.

The ideal version would eliminate these friction points entirely, say a Slack modal that opens directly in your workspace for updates, voice input for those who think better out loud than in writing, and a deeper Slack integration that enables name tagging for special shoutouts. As is, this system works best with teams already comfortable with documentation and it's a significant improvement over Sunday night scrambles.

Progress Check-in Shouldn’t be a Gotcha Moment

"What did you get done this week?" can be the foundation of a culture where everyone sees how their work connects to what matters and shifts the question from “How should we report progress this week?” to “Are we truly making progress?”. 

If you'd like help implementing this, let's chat.