#6fbb44

AI is already inside your company. Maybe not officially. But practically? It’s everywhere.

It’s on every desk, in every browser tab, and in more “quick questions” than leadership would like to admit. People are using AI at home, on their phones, and yes, at work, because it’s fast, helpful, and the path of least resistance.

That creates a risky reality: your team is already experimenting with AI... Just not inside a space you control. And when AI usage happens in the wild, every prompt, pasted paragraph, and uploaded file becomes a potential leak of internal thinking.

To unpack how enterprise teams can adopt AI safely (without freezing up or handing over IP by accident), Arnaud Dasprez, CEO and Founder of HexaGroup, sat down with Igor Carron, CEO and Co-Founder of LightOn, Editor at Nuit Blanche, and Co-Organizer of the Paris Machine Learning Meetup on the Hex-Files Energy Marketing Podcast.

Keep reading for Igor’s clearest takeaways on why AI is different from past tech waves, why data control matters, how “search and reason” changes knowledge work, and what it really means to deploy AI safely. 

Listen to the full podcast here > 

"Now is the time to equip yourselves."

The most common enterprise mistake with AI is waiting for perfect certainty.

Leaders want time to evaluate vendors, draft policies, and “study the landscape.” Meanwhile, employees are already using public tools to do their jobs faster—because they have deadlines, not roadmaps.

And here’s the part that matters: prompts are not harmless.

Every question someone asks contains context, about customers, pricing, internal processes, engineering decisions, commercial strategy, and how your company thinks.

If you don’t provide a controlled internal environment, your people will find their own. And that’s how knowledge leaks quietly — one helpful chat at a time.

Igor’s idea of “equipping yourself” is simple: create a safe space where AI usage can happen inside the walls. That includes:

  • An internal AI workspace that employees can actually access
  • Policies that treat prompts, outputs, and logs as protected assets
  • Controls on what data can be uploaded, stored, or retained
  • Visibility into usage patterns so leaders can spot process gaps

When you control the environment, you’re not just reducing risk. You’re creating trust and laying the groundwork for higher-value use cases.

"What we provide at LightOn is one of the best ways to make sense of your documents.”

If you want the fastest path to enterprise AI ROI, look at your documents.

Most organizations are sitting on mountains of PDFs, scans, specs, drawings, emails, and reports. It’s the hidden nervous system of the business, packed with decisions, standards, lessons learned, and institutional memory.

The problem is that classic enterprise search can’t “read.” It can only match keywords. So it finds documents… but it doesn’t find answers.

AI changes that. Instead of treating a PDF like a dead file, it can extract meaning, structure, and context—so teams can query what’s inside, even across messy formats.

That unlocks practical wins like:

  • Faster access to technical details buried in long reports
  • Smarter connections between related documents across teams
  • Less tribal knowledge dependency (“ask Jeff, he knows where that is”)
  • Better reuse of existing work instead of constantly reinventing it

When AI understands the content, not just the file name, people work more efficiently with greater confidence.

“Finding your document is just the first step.” 

Search has always been step one. The value shows up in step two.

This is where Igor draws a line between basic search and what enterprise teams actually need: search and reason. Not just “here’s the file,” but “here’s the answer—and here’s where it came from.”

In a strong system, AI should be able to:

  • Find the right documents
  • Pull the relevant sections
  • Summarize accurately with citations
  • Handle follow-up questions without 
    restarting the hunt

Think about work like RFP responses, technical proposals, compliance reporting, incident reviews, or maintenance documentation. These tasks aren’t hard because people are unintelligent. They’re hard because the information is scattered, buried, and time-consuming to assemble.

With “search and reason,” the workflow shifts:

  • You ask a question
  • The system finds prior material
  • It drafts a starting point based on real sources
  • Experts refine, validate, and finalize

"The visual cues, I think, are the big differentiator here."

AI used to miss the clues people see at a glance. Small marks, colours, icons, and shapes meant nothing to older models. This created gaps in fields such as engineering, energy, and finance, where visual detail is crucial.

Modern tools read both text and visuals. That shift matters. It means AI can:

  • Read diagrams
  • Understand colour marks
  • Pull meaning from symbols
  • Interpret layout and structure

If AI can't read these cues, it can't give a full answer. When it can, the result feels natural and helpful. This is where multimodal models bring clear value.

“AI setup is never one-size-fits-all.”

Every company asks the same question. Should we run AI on-prem, in a private cloud, or in a mixed setup? Igor’s answer is refreshingly grounded: it depends on your risk profile, budget, and the sensitivity of your data. 

But the guiding principle stays consistent:

Keep the data close. Control the logs. Use computing intelligently.

A practical path often looks like:

  • Keep your data where it already lives.
  • Run the main app near your internal tools.
  • Use GPU power only where needed.
  • Set rules for privacy and log control.

This keeps things safe without slowing progress. That said, the setup is only one piece. Real progress occurs when people utilize the platform. 

Small prototypes inside teams often uncover the best use cases. When staff can test ideas with an internal API, you unlock bottom-up innovation. You also remove fear from the process, because experimentation doesn’t automatically mean risk.

Explore more ideas and practical advice on this topic.

Catch the full conversation with Igor Carron on The HEX-Files, HexaGroup’s energy marketing podcast for leaders who want real results.

Listen Now:

Spotify
Apple Podcasts 
YouTube 

Good to grow? Gauge your readiness in 10 minutes flat.

Take the GOsurvey
BBN

Unlock growth without boundaries

Need region-specific fresh eggs and flying lessons? Learn more about BBN, the agency that unites co-pilots from all corners of the sky. 

Explore BBN