Genealogy Platform
Artificial Intelligence & Data
New York, NY
NeuraFace
Genealogy platforms process huge volumes of user-uploaded images — often historic, scanned photographs that suffer from low resolution, fading, damage or poor lighting — along with inconsistent metadata (names, dates, places). The legacy system was unable to reliably link faces across generations, causing match accuracy to fall and user trust to decline. The client needed a solution that could reliably detect familial resemblance and scale processing rapidly to meet growing upload volumes.
TechBanq developed NeuraFace as a comprehensive, AI-powered facial recognition engine tailored for genealogy and archival imagery. Key capabilities included:
Neural image enhancement & normalization – restoring and standardising old or damaged photographs so they become viable input into recognition pipelines.
Facial landmark detection & alignment – regardless of pose, angle or expressiveness, faces are aligned and normalized for consistent feature extraction.
Deep embedding & similarity engine – converting each face into a high-dimensional vector and using cosine-based similarity to detect kinship patterns between faces across generations.
Metadata-fusion scoring – combining facial similarity with enriched contextual metadata (birth dates, surnames, geographic proximity) to boost confidence in matches.
Scalable architecture & vector indexing – enabling rapid batch processing, real-time lookups and support for millions of simultaneous comparisons.
Impact
NeuraFace re-imagined what genealogy matching could deliver — turning old photographs and partial metadata into meaningful family-link insights. By merging cutting-edge computer-vision with historical context, TechBanq enabled the client to deepen emotional resonance with users while delivering robust, enterprise-grade performance.
The result
Match accuracy improved to ~94%, across old and modern photos.
Average match-processing time dropped from hours to under 5 minutes.
User engagement increased by 60%, driven by higher-quality matches and faster feedback.
The platform regained user trust, reducing dropout and increasing repeat usage.
