I fly mapping jobs out of Springdale, Arkansas. When the batteries are back on the charger, the client still doesn't have a map. They paid for something they can open and send to the next person on the project.
In September 2026 I'm shipping SkyView, the desktop processing and viewer tool on Pilot Ledger. The point is simple. Process drone maps on your own hardware. Do the local photogrammetry on the machine you already own. Then share a link. It is on the waitlist now. It is not in customers' hands today.
The map lives in the wrong place
Cloud processors are easy the first time. You upload the images. You wait. You get a map. You send a link.
Then you do it again next week. A lot of those platforms meter the work. Per map. Per project. Per credit pack. Per reprocess when the client wants one more look at the west pile. You start pricing the job around their meter.
There's another cost that doesn't show up on their invoice. You're uploading raw imagery off a site the client may treat as confidential. A quarry. A plant. A job they haven't announced. The processing sits in a queue you don't control.
Then the share. The quote lives in one place. The map lives in a second portal. The superintendent never bookmarked that login. A week later I get the text. "Where's the ortho?" I hunt for a password I set up under their company name, or I bounce a ZIP through email until it fails.
That's the part of mapping work most software treats as someone else's problem.
What the cloud meter actually costs
I am not going to pretend the big cloud processors are bad tools. They are not. I am not claiming SkyView beats them on features. It doesn't, not yet, and I won't write that.
The pattern is the same. You upload imagery. Processing happens on their servers. You pay per map, per project, per credit pack, or per reprocess. Sharing is strong if the client already lives in that world. If they don't, you walk them into another account.
Cloud processing fees, or a second product to host what you already processed. Data leaving the laptop. A client portal that is one more password.
SkyView's wedge is narrower than a feature war. Process locally. No per-map processing fee. Send a share link. The client does not need a Pilot Ledger login.
Process it on the desk. Share the finished layer.
SkyView is a desktop app. The heavy work stays on your hardware. Orthomosaics, point clouds, DEM and DSM, planimetrics, profiles, sections, GCPs, thermal data, and Gaussian splats. You open them. You inspect them. You process them on the machine in the shop.
Raw imagery does not leave the laptop unless you export something on purpose. That is the rule. Not a cloud-only service that needs the whole stack before it will start.
When a layer is ready, you send the selected layer from Import/Export. Only the deliverable you chose uploads. Not the raw card. Not the whole project. The upload can run in the background while you start the next site.
Files at or under 100 MB go in one step. Larger files split and chunk automatically. I have not put a published point-cloud count cap in the docs. If that changes, I'll say so. I am not going to invent one here.
The web side is a viewer for those shared layers. You get a unique share link. The client opens it. They do not need a Pilot Ledger login. They do not need the desktop app. They need the current ortho, or the cloud, or the surface you actually meant to send.
Your maps. Your hardware. No per-map processing fees.
I'm not going to pretend local work is free
Local photogrammetry uses your CPU, your GPU, your disk, and your afternoon. If the workstation is weak, it will be slow. If the dataset is large, you will feel it. That's honest.
What you don't feel is a new processing bill every time you fly another site for a contractor who already paid you.
I built it this way because that's how I work at Drones Inbound. Quote the job. Fly it. Process it on my own machine. I've been a commercial drone pilot for eight years. The flying got cleaner. The delivery used to get messier as the files got bigger.
The client should not need a scavenger hunt
The other version is a folder named FINAL_v3_USE_THIS, or a processor project under your personal email, or a GeoTIFF that fails in email.
A superintendent on a grading job doesn't want your processing software. They want to open the current ortho, look at the pile, and forward a link to the owner. If that link is a unique share, and they don't have to create an account to see it, they stop asking you where the files went.
I still get the text that says the owner wants last month's ortho next to this month's. I would rather answer that from the same account that quoted the job than from a processor I pay again every time I press go.
What I won't promise
SkyView is not trying to replace your entire geospatial stack. If you need a CAD package or a survey workflow on top of the map, that still happens in the tools you already use. The photogrammetry tools you already own are not going away because I'm shipping a desktop viewer.
SkyView will be included with every paid Pilot Ledger plan when it ships. Pro, Pro + Directory, and Team. No separate SkyView fee on those plans. The public page still shows a waitlist, because it is not shipping today. September 2026.
I am not going to promise a waitlist email I can't verify from here. Go to Pilot Ledger. Join the waitlist. Watch for September.
This is a business problem
Most mapping software talks about GSD and tie points. Those matter. I care about them in the field. The client cares whether the map showed up, whether it is the right site, and whether they can show it to someone else without calling you.
When the processing stays on your desk, and the share is a link, a few things get quieter. You stop losing the delivery in a side account. You stop eating a processing fee that wasn't in the bid.
I still fly. I still process. I still answer the phone when a project manager can't find last week's ortho. I would rather that map come off my hardware than sit in a queue I don't control.
Process locally. Upload the finished layer. Let the client open a share link. It ships September 2026. Join the waitlist on pilotledger.com.