SmileVision reads panoramics, cephalograms, CBCT scans and intra-oral photos the moment you upload them. It numbers the teeth, flags findings and traces landmarks in seconds, and you stay in charge of every call.
One plan · Every studio · Billed in USD
At the end of a long clinic day the eye gets tired. The scan doesn't change. SmileVision reads every image with you, tooth by tooth and landmark by landmark, and shows its working. The final call is always yours.
01 Panoramic X-ray
Drop in an OPG and SmileVision maps the dentition in FDI notation, then marks fillings, crowns, bridges, root canals, caries, periapical lesions, impactions and implants on the exact tooth they belong to.
02 Intra-oral photos
Snap an intra-oral photo and SmileVision outlines calculus, caries, stain, deep fissures and existing fillings. Patients get the picture at a glance. Pair it with an OPG and the findings combine per tooth Beta.
Hover the photo to inspect →
03 Cephalometrics
Upload a lateral ceph and SmileVision places 29 landmarks, draws the tracing and runs Steiner, Downs, Tweed, McNamara, Wits, Jarabak and more against the population norms you choose. Every point stays draggable and every value updates live.
mean landmark error against senior orthodontists on this unseen film · 88 % of landmarks within 2 mm
04 CBCT
4.1 · Segmented
The maxilla, mandible, upper and lower dentition and the mandibular canals are separated automatically and rendered in 3D, right in the browser, in about a minute. Every tooth gets its FDI number too Beta, and you can renumber any of them.
4.2 · Sliced
Scroll axial, coronal and sagittal slices with a 3D localizer, or build a panoramic straight from the volume and send it to OPG analysis.
4.3 · Measured
Check canal proximity and airway volume, measure, plan implants by hand, overlay an intra-oral scan, and export STL or PLY for your lab.
Segmented by SmileVision and rendered live in your browser.
05 Smile Studio
Upload a frontal smile photo and the AI smile design starts by itself: a straighter, brighter smile in seconds, right beside the original. Add a whitening preview, measure the midline, smile arc, cant, gum display and incisor proportions against their published references, and print a one-page smile report for your patient.
It's a simulation to start the conversation, not a promise of the treatment result.
Drag across the photo to compare →
06 Novel Crown
Click the tooth on the pre-op intra-oral scan. The AI finds the tooth and the gum line, and Novel Crown Beta copies the tooth exactly as scanned, hollowed out by the reduction you choose and ending 1–2 mm above the gum. Print it, reduce the tooth to match, and the crown slips on like a sleeve.
The crown's shape is computed from your scan, not generated by AI. Every design comes with its own checks, such as wall thickness and retention. It is in beta, so check the fit before clinical use. Already prepared? A second mode designs the temporary from the preparation scan.
Drag the model to turn it →
07 The platform
Every screenshot below is SmileVision itself, running the same AI on the same public images you just scrolled through.
08 Always learning
Every correction you make, with your consent and de-identified, becomes a lesson. New models train overnight, sit an exam on a frozen test set they have never seen, and ship only if they beat the model you're using. Then the loop starts again.
Example: our latest ceph release
Released only after beating the previous model on held-out films. Nothing ships on a hunch.
09 Founders
SmileVision was built at the chair and at the keyboard at the same time. One founder decides what a clinician needs to see; the other makes the machine see it.
Co-founder · TechnologyAI & engineering
Zauraiz leads SmileVision's AI and engineering. He holds an M.Phil in Engineering specialising in artificial intelligence and an honours degree in Mechatronics. He has spent a decade shipping AI, web and blockchain products, from peer-reviewed computer-vision research at MIMOS Berhad in Kuala Lumpur to AI for dental communication platforms and the CTO seat at Vitalum. At SmileVision he builds the models, the overnight training loop and the platform around them.
Co-founder · ClinicalSpecialist orthodontist
Dr Putu completed his orthodontic specialist training at Universitas Airlangga in 2016. He practises in Surabaya at D-Arts Dental Clinic and is a clinical advisor to Smile Cloud Dental Clinic, focusing on temporary anchorage devices, clear aligners and digitally planned, predictable treatment. He has presented case studies at an Asia-Pacific orthodontic specialist forum. At SmileVision he sets the clinical bar, from report formats clinicians actually use to reviewing what the AI gets right and wrong.
10 Pricing
No demo call before you see a price, and no tiers to decode. You get the whole platform, and every model we ship after you join.
Prices in USD; local taxes may apply. Already a member? Log in
No. SmileVision is decision support. It points out what it sees and shows where, and you confirm, correct or dismiss every finding. The clinical decision always stays with the clinician.
Panoramic, bitewing and intra-oral images as JPG, PNG or BMP up to 30 MB. Cephalograms also as TIFF or DICOM. CBCT as a zipped DICOM series. Smile photos as JPG, PNG, WebP or HEIC, and intra-oral scans for Novel Crown as STL, PLY or OBJ.
Annotated images, an FDI tooth chart, PDF reports, a one-page Steiner and Downs ceph report, ortho records reports, a one-page smile report with the before-and-after, STL files for Novel Crown, STL or PLY from CBCT Studio, and one master report per patient from Fusion Studio.
Only if you opt in. When you share an image to help improve the AI, patient names, IDs and file details are removed, and printed text on X-rays is blurred during review. You can stop sharing for your clinic at any time.
We measure every model on held-out images it has never seen, and publish a new one only when it beats the current model. For example, our ceph AI's mean landmark error is about 1 mm on held-out films. AI can be wrong, which is why every result is shown with its evidence and stays editable.
Two co-founders: Zauraiz Alamgeer, an AI engineer, and drg. Putu Indra, Sp.Ort, a specialist orthodontist. Clinical feedback from practising dentists shapes every release.