Enterprise leaders are dealing with a well-recognized problem at an unfamiliar scale.
Each group is being requested to maneuver sooner as markets change rapidly, buyer expectations proceed to rise, and know-how advances at a tempo that may really feel overwhelming. Groups are anticipated to ship larger outcomes, typically with the identical assets they’d earlier than.
AI helps organizations meet these expectations. A few of the strongest examples I’ve seen revolve round scaling the judgment, technique, and success measures that robust performers already set for themselves and their groups. AI brokers apply that experience constantly throughout a rising quantity of labor, serving to them ship extra with out sacrificing high quality.
We’ve seen it firsthand on my crew. As innovation cycles have accelerated, product launches have elevated from a quarterly cadence to weekly—and generally even every day—occasions. Our groups at the moment are supporting a rising quantity of launches, as much as 150% 12 months over 12 months.
To alleviate the strain, we’ve appeared for locations the place AI may help groups at Microsoft discover the correct data sooner, cut back repetitive coordination, and convey extra consistency to work that is determined by shared context. To do this, we used Microsoft Foundry, Microsoft’s platform for constructing and managing enterprise AI functions, to create brokers grounded in enterprise data and embedded within the stream of labor, serving to our groups function at larger scale whereas staying targeted on the work the place their experience issues most.
Why context issues
One lesson turned clear in a short time: AI is barely nearly as good as the information it has entry to. Normal-purpose AI can generate content material, however enterprise selections depend upon data unfold throughout paperwork, workflows, enterprise programs, communications, and institutional data.
For us, Microsoft IQ helped join that enterprise context to our AI capabilities. Moderately than asking workers to assemble data from a number of sources, brokers may draw from the identical data individuals depend on on daily basis to floor related data and help higher selections.
However IQ does greater than floor AI in the correct knowledge. It helps join the data and workflows that form how the enterprise really operates.
That shift modified the function AI may play. As a substitute of merely serving to individuals discover data, it may assist groups work from a shared understanding of what’s occurring throughout the enterprise.
Context alone wasn’t sufficient. The breakthrough wasn’t a single agent. It was making a approach for groups to construct on what was already working.
As individuals shared profitable brokers and AI abilities, experience began changing into simpler to reuse and scale. Concepts that started with one crew may rapidly create worth for a lot of others.
Microsoft Foundry turned necessary as a result of it allowed us to floor brokers in organizational data, join them to current workflows, and operationalize them past a single crew.
In some ways, this displays a broader lesson throughout AI adoption. As Jay Parikh just lately wrote, “AI alone doesn’t rework a enterprise. The system round it does.” The next examples present what that appeared like inside our advertising group:
Elevating the standard bar at scale
As our Microsoft Foundry enterprise grew, so did the quantity of content material we wanted to create. Our crew now evaluations and publishes greater than 200 weblog posts annually, sustaining a constant high quality bar more and more depending on a small variety of material consultants. A lot of their time was spent making use of the identical assessment standards again and again.
Moderately than reviewing each draft from scratch, one in all our content material leaders documented the rubric she makes use of to judge a powerful weblog and refined it till it mirrored the requirements our crew anticipated.
Utilizing Microsoft Foundry, we translated that expert-defined rubric right into a repeatable workflow that would determine gaps and alternatives earlier than content material reached a human reviewer. The aptitude was built-in instantly into the content material creation course of, bringing on the spot suggestions to each drafted publish and making expert-defined requirements accessible to each content material creator.
Evaluate cycles that after required substantial handbook effort can now be accomplished in minutes, leading to increased satisfaction and over 2,000 estimated hours saved yearly throughout the crew. Extra importantly, the method demonstrates a broader sample organizations can apply in lots of domains: use AI to use established standards at scale so consultants can focus their time the place judgment, teaching, and expertise create essentially the most worth.
What we automated is consistency, not judgment. Our crew set the bar primarily based on our experience; the AI agent evaluations each publish in opposition to that bar.
Validating messaging earlier than it reaches clients
Because the tempo of innovation accelerated, one query stored developing: would our messaging resonate with the purchasers we have been attempting to succeed in?
At Microsoft, we goal to maintain the shopper on the middle of every little thing we do. That led us to search for methods to judge messaging earlier than it reached clients, utilizing greater than inside opinions alone.
We utilized that method via AI Messaging Assistant (AMA), which helps consider messaging and positioning in opposition to totally different viewers views earlier than going to market. As a substitute of relying solely on inside opinions, groups can pressure-test whether or not a message is evident, related, and actionable for the stakeholders they’re attempting to succeed in.
Utilizing Microsoft Foundry, we grounded AMA with a digital congress of personas primarily based on actual buyer conversations and prolonged it with the experience, product data, messaging steerage, and enterprise context our groups depend on on daily basis. That made it attainable to maneuver from a one-off AI experiment to a repeatable workflow the place groups may consider messaging in opposition to a shared understanding of viewers wants relatively than rebuilding that understanding for each assessment.
The broader sample is utilizing AI to pressure-test necessary selections earlier than they attain clients, companions, or workers.
Protecting groups aligned because the tempo accelerates
As launch exercise accelerated throughout our enterprise, conserving groups aligned turned more durable than creating the work itself. New bulletins arrived every day. Priorities shifted rapidly. Data was unfold throughout planning backlogs, documentation, conferences, and operational programs. Our entrepreneurs have been spending an excessive amount of time assembling context and never sufficient time performing on it.
In response, we began by writing down how our advertising work really will get accomplished, turning an unwritten course of into a transparent specification. With that in hand, we may kind the work: which elements required a marketer’s judgment, which may very well be automated, and which may very well be delegated to AI.
Utilizing brokers constructed on Microsoft Foundry, we related the programs our groups already depend on, together with planning backlogs, documentation, assembly indicators, and different operational sources. Moderately than manually gathering updates from dozens of locations, groups can work from a real-time view of key developments, upcoming launches, and modifications that have an effect on go-to-market plans.
This reworked alignment from a handbook effort right into a repeatable workflow. As a substitute of spending time assembling data, groups spend extra time understanding what modified, why it issues, and what actions to take subsequent.
The problem was by no means a lack of awareness. It was coordinating that experience throughout a quickly altering setting.
The result is just not merely sooner communication. It’s higher organizational alignment. When groups function from the identical base, selections occur sooner, handoffs turn out to be smoother, and organizations can reply extra rapidly to vary.
Scaling experience, decreasing friction
Throughout every of those examples, the objective wasn’t automation for its personal sake. The objective was making experience accessible wherever it may create worth. Trying again, the lesson wasn’t {that a} single AI functionality modified how we labored. It was that constructing the correct system round these capabilities allowed experience, context, and judgment to scale throughout the crew.
Know-how will proceed to evolve. The tempo of enterprise will proceed to speed up. However the differentiator stays the identical: Folks present the judgment. Folks set the technique. Folks outline success. AI helps them scale it.
