Скачать или смотреть What's a Knowledge Graph? Using Linkedin as an Example, Let's Talk it Through.

Knowledge graphs and as a result, graph databases, are a big part of what excites me about AI agents and compound AI systems. Let me explain why.

Oversimplifying it but a Knowledge Graph is a way to understand the interconnected nature of all of your company’s data sets.

The way data relates to other data and therefore where there might be hidden opportunities to capitalize on or risks to mitigate. The kinds of things that make you say “wow, I did not know that…” out loud, even when no one is around… ;)

Let’s use LinkedIn as an example.

It’s one thing to understand how many connections a person has on Linkedin but, when viewed through the lens of a CRO, it’s an entirely different thing to understand where each of those connections work, the strength of the 1st degree connections, the associated 2nd degree connections of those people and how the collective network does or doesn’t align with tier 1 target companies, executive buying committees, how accessible those connections are / aren't etc. that the CRO is looking to cultivate relationships with.

Anyhooooo, you get the idea.

Now when you combine Knowledge Graphs with Generative AI, things get interesting, quickly.

Two examples that come to mind.

Next Best Action:
What if you built an AI agent that could comb through a proprietary, internal graph database of your top 1000 clients and uncover hidden insights invisible to the Sales team because, understandably, they can’t invest that kind of time to analyze that amount of data…

Then the AI agent could cross-reference the data with real-time market developments and your company’s content library to then, in conjunction with an LLM, come up with a personalized, relevant 1:1 sequenced outreach campaign to 250 executives with a projected open rate of 44% and a conversion rate of 5% at an average deal size of $150K based on historical campaigns?

Added bonus, both the 5% and 95% would also feel like you added value to their world vs the “Just checking in?” email that doesn’t go over so well due to the thoughtful personalization…

Product Innovation:
Imagine an AI Agent that maps out the universe of your complex customer relationships including every customer journey map, touchpoint, interactions, product usage data, trailing 6+ months of call transcripts, risk profile, upcoming life events etc to identify patterns, preferences, clustering of insights, ideas and surface areas of opportunity for new features, incremental revenue etc?

Enabling next level human + machine product innovation.

I’m 110% here for that Digital Product Innovation session (s/o DPI Rockstars: Marco Ambrosio, Barry Fiske, Greg Boullin).

I’m 110% here for the Knowledge Graphs.