A small file hunt became a bigger idea

The file was somewhere.
That was the problem.

My phone. My Downloads folder. My computer. Dropbox. Somewhere inside an AI workspace.

I was ready to get some work done. First, a little treasure hunt.

Find. Download. Upload. Again.

I work with a lot of material: research, reports, business plans, financial models. The useful work was in those files. Getting the right ones into the right AI tool had become a job of its own.

I’d find a document, download it, upload it somewhere else, then repeat the process for another conversation or another person. Sometimes the next destination was simply a different AI.

My files were doing a remarkable amount of travelling. I was spending too much time arranging the trips.

Every extra transfer slowed the work down. It also left another place to look the next time I needed something.

Then there was the part I couldn’t see.

Once I’d handed over the material, I had no clear record of how the AI systems accessed it. Which file had been requested? Which part? What had the service actually prepared in response?

I wanted my files to be easier to find. I also wanted that paper trail.

What if the files had their own home?

A place I could return to. Choose a pack. Connect the AI I wanted to use. Give someone else access when it helped. Decide what they could consult, and for how long.

Changing AI tools shouldn’t mean starting another file hunt.

That became OpenDataRoom: your data room, your choice of AI. A simple way to bring an authorised AI connection to a set of files, with a record of the requests it makes.

I wanted that independence for myself, too.

My work often brings together research, business plans, financials and months of careful thinking. I want to decide who sees that material and what they can do with it. Choosing a useful tool should leave those decisions with me.

So I wanted OpenDataRoom to be a neutral third party: a place to organise my packs, control access and keep a useful record. I have no interest in turning someone else’s private work into our research, an investment opportunity or a dataset.

Our commitment is straightforward: your files and private request details are used only to provide and protect the service you choose. We do not sell them, train AI models on them, or use them for our own research, investing or profiling.

You choose the material, the people, the access window and the compatible AI. That AI receives the content it retrieves within your approved scope and has its own data policies. Our job is to make your choices clear and easy to manage.

The idea kept finding useful work.

I prepared a hotel research pack for a marketing consultant. I brought deep research, audits, calculation checks and a startup’s business plan and financials together for an investor I advise. I used the same approach for an investment committee deck.

For conferences and events, I created research packs that fundraising teams could consult through their own AI. They could bring the questions that mattered to them, with the supporting material already assembled.

Each handover made the idea more useful: a coherent pack, deliberate access, a clearer view of the requests coming back. A search that finds nothing can also give me something to investigate before improving the material.

The files are ready when I am.

Now I have my data packs and a straightforward way to access and use them wherever I’m working. I decide what to share, the security policies around it and who gets permission.

I can get straight to the useful part: asking a better question, testing an idea, comparing options or building on work I have already done. Once connected, I can return to approved packs without repeating the whole setup. Less file hunting. More momentum.

Keeping the source material, analysis and earlier findings together helps me get more from the AI tools I choose. I can check that the right information is available, investigate gaps and carry the work forward without rebuilding the context each time. The AI has evidence to consult; I have sources to check its answers against.

It gives me a data room I can trust—and the control I need to share with confidence.

If you’ve ever searched three folders and two apps for a file you were absolutely certain you had, this started with you in mind too.