Lightroom does 150.
ShotSelect is the keyboard-first, on-device AI culling layer for photographers — faster by design, offline by default, no subscription required to start.
Photographers lose a full workday
before editing a single image.
A wedding photographer shoots 3,000 frames per event. Lightroom averages 150–200 photos/hr. That's 15–20 hours of post-production before the real work starts.
Import → render → click flag → move to next. Workflow unchanged since 2007. Maximum 200 photos/hr by design.
3,000 photos = 3,000 mouse clicks. No professional culling app is keyboard-native. That single shift would halve session time.
Aftershoot and Narrative Select need internet, accounts, and subscriptions. Useless on location or behind venue Wi-Fi blocks.
Lightroom + Aftershoot + Narrative = $420–1,956/yr. After Adobe's 2024 price hike, migration intent is at a decade high.
AI in the session.
Human in the loop.
ShotSelect is a dedicated keyboard-first culling session with six on-device AI models running in parallel. AI surfaces signals on every frame. The photographer decides. The keyboard executes.
Arrow up to keep, down to reject — auto-sorted in real time. No mouse. No clicks. 1,000+ photos/hr on any hardware.
Blur detection, duplicate clustering, closed-eye flagging, face grouping by person, emotion scoring, and YOLO-World object search — all running on CPU, fully offline.
Type "first kiss" or "bride laughing" — on-device semantic embeddings find matching frames across thousands of photos. No cloud. No index upload. Instant.
Ratings, picks, and labels write as XMP sidecars next to RAW files. Open in Lightroom Classic, Capture One, or Bridge — everything syncs automatically.
17 million photographers.
Zero dedicated culling tools.
Photography software is a $1.2Bn market dominated by editing suites. No company owns the culling layer. The dedicated AI culling category is entirely unclaimed.
- ›Adobe treats culling as a footnote inside a $660/yr editing suite
- ›Photo Mechanic solves ingestion speed — no AI, no keyboard-native flow
- ›Aftershoot + Narrative built AI for selection suggestions, not the session itself
- ›No tool ships: keyboard-first + on-device AI + offline + no login
Every competitor has a gap.
We sit at the intersection.
| Lightroom | Photo Mechanic | Aftershoot | Narrative Select | ShotSelect | |
|---|---|---|---|---|---|
| Keyboard-first culling | Partial | Yes | No | Partial | Yes |
| On-device AI · fully offline | NoCloud only | No AI | PartialNeeds cloud account | PartialNeeds activation | Yes100% offline |
| Natural language search | No | No | No | No | YesOn-device semantic AI |
| No login required | No | Yes | No | No | Yes |
| AI informs — human decides | — | — | NoAI automates cull | Partial | Yes |
| Pricing | $120–660/yr | $299 one-time | $180–576/yr | $120–720/yr | Free → $99/yrCheapest with AI |
Product complete.
Beta live. Paid tier next.
ShotSelect v1.0.18 ships on macOS today with automated license delivery, full CI/CD, and all 6 AI models in active use by beta photographers.
- ✓Full keyboard culling — arrow keys, tag slots 1–9, compare mode
- ✓6 on-device AI models with background task queue
- ✓Natural language semantic search (on-device embeddings)
- ✓XMP sidecar export — Lightroom + Capture One compatible
- ✓SQLite session persistence — resume across restarts
- ✓Ed25519 offline license system — cryptographic, no server
- ›Paid tier live — Pro $99/yr hard paywall trial
- ›Mac App Store submission + Windows public beta
- ›Client review links — share curated gallery with clients (beta)
- ›AI-suggested culling — one-key accept on model picks
Free to start.
$99/yr to stay professional.
Hard paywall trials convert 5× better than freemium (10.7% vs 2.1%, RevenueCat 2024). Full-featured 14-day trial → clear Pro paywall. Lifetime option anchors the price for subscription-averse photographers.
- Keyboard culling session
- All 6 AI models
- XMP export · 40+ RAW
- All AI models · unlimited
- Session history · unlimited
- All future model updates
- Mac + Windows · one license
- Everything in Pro
- Shared sessions (team)
- Client review links
- Priority support
Target vs 10.7% industry median · RevenueCat 2024
Path to $1.5M ARR
in 24 months.
No paid acquisition in Year 1. App Store + photography community distribution. 80%+ gross margin at scale.
Lifetime licences modelled separately: est. 2K @ $199 = $398K one-time by FY27 end, not in ARR above
Technical founder.
Full stack, day one.
ShotSelect was designed, built, and shipped solo — from Electron architecture and on-device ML inference to CI/CD, code-signing, and beta distribution infrastructure.
Built ShotSelect end-to-end: Electron + React + TypeScript desktop app, six ONNX on-device AI models (YOLOv8, InsightFace, semantic embeddings), Ed25519 offline licensing, and full CI/CD pipeline (GitHub Actions, Cloudflare R2, Apple notarization). Shipped v1.0.18 across Mac and Windows.
- ✓Electron + React + TypeScript app (Mac + Windows)
- ✓6 ONNX on-device AI models + semantic search
- ✓Offline Ed25519 license system via Cloudflare
- ✓CI/CD: build → sign → notarize → distribute, automated
- ›Head of Growth — photography community (WPPI, YouTube, Reddit)
- ›ML Engineer — AI model accuracy + new detection capabilities
- ›iOS Engineer — mobile companion (Studio tier upsell)
Pre-seed round.
One milestone: $1M ARR.
Post-product, pre-revenue. The product is complete. Funding buys distribution infrastructure and the first two hires.