Is your AI working group feeling lost?

Here’s how to get back on track

It often starts with excitement and urgency.

LPs are asking about your AI strategy, peers are talking about automation, and tools promising to revolutionise workflows seem to launch every week.

So you form a working group. You trial vendors, pay for proofs-of-concept, and experiment with new tools. But the results don’t always match the hype. Some use cases show promise, while others fizzle. Adoption is patchy. Teams lose confidence. Before long, the question becomes: is it really worth pursuing an AI strategy at all?

According to Filament Syfter’s Josh Berman and Martin Pomeroy in a recent discussion with the-line.net, most firms they speak to are in exactly this position. 

The key is not to give up, but to reset.

In Josh’s recent article for theline.com , he outlines seven ways to help your AI working group refocus and prepare for what’s next: agentic AI.

1. Start with the problem, not the tool

One of the biggest missteps is letting technology dictate the conversation. Instead, define the business problem you’re trying to solve; deal origination, portfolio monitoring, internal knowledge sharing - and build from there.

2. Build your “fuel” before your engine AI is only as effective as the data behind it.

Without clean, connected, and accessible data foundations, even the best tools will fail to deliver.

3. Focus on behaviour change, not just tech

Adoption is the toughest challenge. If AI adds friction, users won’t engage. The goal is to make new workflows seamless — even invisible.

4. Act like a product team

The most effective groups think like product managers, asking what users really need, where workflows break down, and how to create better experiences.

5. Prepare for agentic AI

We’re moving beyond “ask and answer” towards agents that complete tasks end-to-end — pulling data, building outputs, uploading to systems, and adapting along the way. This shift will change the role of humans in the process.

6. Don’t just build for efficiency, build for advantage

Shaving minutes off existing tasks is useful, but not transformative. The real edge comes from insights and capabilities competitors can’t access.

7. Stop chasing features and start building strategy

Most vendors offer the same features to everyone. What sets firms apart is a roadmap grounded in specific business needs, strong data foundations, and the ability to evolve.

The bottom line

The messy middle is a natural stage in the AI journey. What matters now is how you move forward. Success in the age of agentic AI won’t come from the flashiest demos, but from the clearest strategies.

Register with the-line.net to read the full article by Josh Berman: Is your AI working group feeling lost?

Find out more about how the data and tools within Syfter can make a huge difference to your deal origination, relationship building and data integration with your own systems, with a personalised demo, today.