For years the mantra across the web was "content is king." With AI now woven into how organizations work, I'd argue a new king has emerged: context. It's a theme I believe was at play well before these new technologies: having the right context at the right time is critical to keeping your business moving forward and making the right decisions.
What do we mean by context? Clients come to us for solutions to a problem they're facing or a system they're trying to build. The first step is often understanding the perspective they're coming from, and that takes skills learned over many years. For us, understanding the client's business, their pain points, what's changing in their business or industry, and their technology stack are all core to how we advise on solutions. This is the context that drives the right solution. For years, this context has lived in documents, emails, project management systems, or, worse yet, in a single person's head on our team.
With the advent of conversational AI systems, we can create a shared, secure context layer for our entire team and our clients that is easily indexed, updated, and queried. We're finding this to be a killer feature: it continuously improves the accuracy of our AI systems and gets the right context to the right team members at the right time.
This ties into what we've previously written about regarding what is and isn't working in AI implementations. When we give a team or employee access to a powerful LLM, it brings a mountain of knowledge, but in many cases it lacks the specific context that person needs in the moment. Take questions like "Who is the decision maker on the Wing Ding account?" or "What module in our ERP handles landed costs?" These questions are unique to your organization, and the model most likely won't have that information.
So how are we using this at Sego Solutions? On the client-facing side, we built Sego.pm, our custom tool for shaping client requests against the project's requirements and current state. It lets us scope tasks in far more impactful and productive ways than we ever could before. We drew on many of the principles from Shape Up, the popular book from 37signals, and we'll be writing more about Sego.pm in a future post.
We're also deploying shared context within organizations using memory layers such as Garry Tan's open-source gbrain. This lets us deploy context in a way that doesn't lock you into a single AI vendor and keeps your company's knowledge yours. We're still learning a ton as we use these systems, and they're giving us an exciting glimpse into how organizations will provide the right context at the right time to improve the speed, accuracy, and quality of their decision-making.
Here's a good exercise: think about the information in your organization that requires digging through your inbox, pinging a co-worker, or clicking around a legacy application to find. That's your context gap, and it's exactly what an AI system can't help with until you close it.
If any of these topics interest you for your organization, we'd love to chat. Reach out any time!