An Assistant Connected to Every Tool in Your Workflow

A helper that lives off to the side of your actual tools is only ever going to be a novelty. Ask it something, get an answer, then go do the real work somewhere else. Imagen’s assistant was built differently: it sits inside the same project as your culling, editing, and export, connected to the tools you already use, so a request does not dead-end in a text box. You ask, it finds, and then you can act immediately without switching anywhere.

One assistant, wired into the whole project

Because the assistant lives inside your project rather than beside it, a search is never the end of the interaction. Ask for a specific set of photos and the results appear inside the same gallery you are already working in, ready to be rated, edited, or exported without any extra step to move them somewhere else first.

That connection is what separates a useful assistant from a search bar with a friendly tone. The value is not the conversation itself, it is what happens immediately after the conversation, inside the same space where the rest of your work already lives.

Finding photos in your own words

Instead of relying on keywords you tagged ahead of time, you describe what you want the way you would describe it to another person. Pull up every shot from the ceremony processional. Surface anything with the two groomsmen who kept photobombing the formal shots. Find the handful of frames where the light through the window actually worked. Each of those is a plain description, not a search operator, and the assistant works from that description directly.

This matters most on large galleries, where scrolling to find a mental image of “that one shot” can eat far more time than describing it in a sentence and letting the assistant do the searching.

Acting on what it finds, immediately

Once the assistant returns a set of photos, the next step is right there. Rate the set, apply an edit, export just those frames, or hand them off to another part of your review, all without leaving the flow you were already in. There is no export-and-reimport step, no separate app to switch to, and no copying file names between two different tools.

That immediacy is the actual time savings. A smart search that returns a list you then have to go re-find manually in a different screen barely saves any time at all. A search that lands you directly on editable, exportable photos saves quite a lot.

You steer, the assistant supports

None of this is meant to run on autopilot. The assistant surfaces and helps you act, but every decision about what stays, what gets edited, and what ships to a client is still yours. Think of it as control on demand and automation on demand, offered as choices rather than defaults you are stuck with.

The Imagen assistant finding and editing photos from a plain-language request

You can lean on the assistant heavily for a large, repetitive gallery, or barely use it at all on a smaller job where you would rather work by hand. Both are legitimate ways to use the same project, and neither one is the “correct” way to work.

Where it fits next to manual review

The assistant does not replace rating, flagging, or manually scanning a gallery, it sits alongside those habits. Many photographers will keep doing a manual first pass out of habit and use the assistant for a targeted follow-up search, like pulling every photo of a specific guest after the fact for a special request. Others will lean on it earlier, using it to triage a huge gallery before doing any manual work at all. Neither approach is more correct than the other.

A practical example

Picture a corporate event with a few thousand frames and a client who wants a highlight reel focused on three specific executives. Scrolling through the entire shoot looking for those three people would eat an afternoon. Describing them to the assistant and asking it to surface every frame they appear in turns that afternoon into a much shorter task, leaving you time to actually edit the highlights instead of just finding them.

The same pattern holds for smaller requests too, like a bride asking a week later for every photo that includes her grandmother. Instead of reopening the entire gallery and scrolling from the start, you describe the request and get a working set back in moments.

Still early, and still growing

This is a first version of a connected assistant, not a finished one. Expect the range of requests it handles well to keep expanding over time, without a fixed date attached to any specific addition. What will not change is the underlying idea: the assistant stays wired into your tools rather than sitting apart from them, so anything it finds is immediately something you can act on. For more on how the pieces of the new Studio work together, our valuable tips library has further reading.

Search that understands your gallery’s structure

A generic search box treats your photos as a flat pile of files with names and dates. The assistant works with a much richer sense of what is actually inside your project: which frames belong to which part of a shoot, which ones already carry a rating or a smart tag, and which ones sit next to each other in the timeline. That context is part of why a plain-language request works as well as it does, since the assistant is not just matching words, it is reasoning about a structured project rather than a folder of anonymous files.

This matters in practice whenever a request depends on more than visual content alone. Asking for photos from a specific part of the day, or combining a description with an existing rating or tag, works because the assistant can see all of that structure at once rather than only the pixels in each frame.

Why disconnected search tools fall short

Plenty of tools can search a folder of images by keyword or metadata, and some can even describe what is in a photo reasonably well. What most of them cannot do is hand you a result you can immediately act on inside your actual editing workflow. A result that lives in one app while your edits live in another means copying files, renaming things to keep track of a selection, or exporting a list and reimporting it somewhere else.

Every one of those extra steps is a place where time leaks out of a process that should be simple: describe what you want, get it, use it. Keeping the assistant inside the same project as everything else is what removes those extra steps entirely, rather than just making the search itself a little smarter.

Building it into your regular workflow

The assistant earns its place fastest when it becomes a habit rather than a novelty you try once. A natural point to reach for it is right after a big import, to get an early read on a gallery before you have rated anything, or near the end of a project, when a client asks for something specific that your existing ratings and tags do not cleanly capture. Over a few projects, most photographers settle into a pattern of when the assistant genuinely saves time versus when a quick manual scroll is just as fast, and both are legitimate calls depending on the shoot. There is no wrong time to use it, only patterns that end up feeling more natural than others the more projects you run through it.

Why the core matters more than the edges right now

Early versions of any assistant feature tend to get judged on their edges rather than their core, and that is a fair way to judge them. The core, right now, is dependable plain-language search connected directly to the rest of your project, and that core is solid enough to build a real habit around even while the edges keep expanding. Both examples earlier in this article share the same shape: a specific, describable request that would otherwise mean scrolling through a gallery from the beginning.

What’s new in Imagen: Here is the short version: Imagen is now an intelligent, agentic Studio for professional photographers, with culling, editing, review and export in one project. The assistant lives inside that project and is connected to Imagen’s tools, so a plain language request can find photos and then act on them, with you steering every decision. The Studio is rolling out gradually as a beta.

Frequently asked questions

Do I have to leave my project to use the assistant?

No. It works inside the same project as your culling, editing, and export, so results appear where you are already working rather than in a separate tool.

Can I act on the photos the assistant finds right away?

Yes. Once it returns a set of photos, you can rate, edit, or export them directly, without an extra step to move them somewhere else first.

Does using the assistant mean I give up manual control?

Not at all. Every decision about what stays and what gets edited is still yours. The assistant supports that process, it does not take it over.

Will the assistant handle more complex requests over time?

That is the direction it is headed in. Capability is expected to keep expanding, though no specific dates are attached to future additions.

Ready to try an assistant that’s actually wired into your workflow? Start a project in Imagen’s new Studio, where you cull, edit and export together, at imagen-ai.com.