To build coordinated AI agent systems, firms must first extract siloed operational knowledge. This involves not just digitizing documents but systematically observing employee actions like browser clicks and phone calls to capture unwritten processes, turning this tacit knowledge into usable context for AI.
Related Insights
Instead of merely 'sprinkling' AI into existing systems for marginal gains, the transformative approach is to build an AI co-pilot that anticipates and automates a user's entire workflow. This turns the individual, not the software, into the platform, fundamentally changing their operational capacity.
To successfully automate complex workflows with AI, product teams must go beyond traditional discovery. A "forward-deployed PM" works on-site with customers, directly observing workflows and tweaking AI parameters like context windows and embeddings in real-time to achieve flawless automation.
Effective enterprise AI needs a contextual layer—an 'InstaBrain'—that codifies tribal knowledge. Critically, this memory must be editable, allowing the system to prune old context and prioritize new directives, just as a human team would shift focus from revenue growth one quarter to margin protection the next.
Counterintuitively, the path to full automation isn't just analyzing conversation transcripts. Cresta's CEO found that you must first observe and instrument what human agents are doing on their desktops—navigating legacy systems and UIs—to truly understand and automate the complete workflow.
The effectiveness of agentic AI in complex domains like IT Ops hinges on "context engineering." This involves strategically selecting the right data (logs, metrics) to feed the LLM, preventing garbage-in-garbage-out, reducing costs, and avoiding hallucinations for precise, reliable answers.
Run HR, finance, and legal using AI agents that operate based on codified rules. This creates an autonomous back office where human intervention is only required for exceptions, not routine patterns. The mantra is: "patterns deserve code, exceptions deserve people."
The durable investment opportunities in agentic AI tooling fall into three categories that will persist across model generations. These are: 1) connecting agents to data for better context, 2) orchestrating and coordinating parallel agents, and 3) providing observability and monitoring to debug inevitable failures.
Leadership often imposes AI automation on processes without understanding the nuances. The employees executing daily tasks are best positioned to identify high-impact opportunities. A bottom-up approach ensures AI solves real problems and delivers meaningful impact, avoiding top-down miscalculations.
AI tools like LLMs thrive on large, structured datasets. In manufacturing, critical information is often unstructured 'tribal knowledge' in workers' heads. Dirac’s strategy is to first build a software layer that captures and organizes this human expertise, creating the necessary context for AI to then analyze and add value.
The ultimate value of AI will be its ability to act as a long-term corporate memory. By feeding it historical data—ICPs, past experiments, key decisions, and customer feedback—companies can create a queryable "brain" that dramatically accelerates onboarding and institutional knowledge transfer.
