The Paradox of Planning: Why your Enterprise AI Strategy is putting you behind
Every week I speak to half a dozen executives who have an AI task force and are trying to define an AI strategy.
Teams exist in the classic build vs buy dilemma.
They have pain they know AI can solve — and often think they are better served by a bespoke solution rather than an off the shelf solution.
They want to evaluate every possible solution before committing.
They spend time debating, writing policies — and often take months before doing a real experiment that puts AI in action.
Often the executive hours spent on the contemplation cost more than the AI experiment itself.
Here are 3 pieces of advice I give leaders who are thinking about how to approach AI:
1 — AI is as valuable as the corpus you feed it
Garbage in, garbage out.
Teams who think about building their own LLM solutions often don’t consider how they will make their long, complicated documents readable by LLMs. If you don’t get this step right the rest of it doesn’t matter.
Ingesting complex, different document types from different sources is critical to uncovering the information and insight you’re actually looking for.
2 — Find strategic partners
Developing AI solutions is most effective when done with expert partners, with specialized AI knowledge and resources.
Strategic partners allow for:
- Accelerated time to market (and opportunities for faster prototyping)
- Scalable infrastructure and co-development opportunities
- Access to latest research and analytics capabilities
- Reduced risk and costs
3 — If you haven’t started — you are already late
Early adopters of AI solutions have already yielded exponential workflow gains, compounding as they continue to scale across credit, equity and research teams.
The funds who have started to deploy AI continue to get further ahead of those still contemplating, or designing internal experiments.
With Hebbia, they have:
- An LLM readable data layer, including a custom metadata layer
- An Enterprise grade web app that plugs directly into workflows, and enables AI based collaboration across teams as they scale
- Access to APIs for collaborative experimentation and prototyping
- The ability to give models feedback for ongoing improvement and customization
- + the right 1:1 onboarding that enables users to get the most out of AI’s capabilities (reliably, and accurately)
To the Enterprises who are spending more time planning their AI strategy:
Planning won’t put you further ahead — getting started will.
This post was written by Lainie — leading Operations and Product at Hebbia.