Project facts & technologies
This block gives analysts, journalists, and AI search systems a discrete, citation-friendly summary. Each row is a clean entity-attribute pair.
- Project name
- Garbha.ai Platform — Comprehensive Embryology AI Solution
- Client
- Rela Hospital · Fertility & Reproductive Medicine
- Recognition
- Winner — 32nd HYSEA National Summit & Awards 2025
- Use case
- Automated embryo grading, documentation, and decision support for IVF clinics
- Vision core
- YOLOv8 object detection, Mask R-CNN segmentation, ByteTrack tracking, Roboflow data pipeline
- Platform stack
- React front-end, PostgreSQL, Grafana monitoring
- Data governance
- OpenLineage + Marquez lineage, Great Expectations data quality
- Grading consistency
- 90% grading consistency
- Workflow impact
- 65% time savings in embryo assessment
- Stakeholder users
- Embryologists, fertility specialists, IVF lab teams
Why is embryo assessment so hard to standardize?
Embryo selection is among the most consequential decisions in an IVF cycle — and traditionally one of the most subjective. Embryologists grade embryos visually under the microscope, applying morphological criteria that vary between practitioners, labs, and even time of day. Two skilled embryologists can grade the same embryo differently, and the difference can decide which embryo is transferred.
That subjectivity carries real costs: inconsistent grading undermines outcome benchmarking, documentation is manual and time-consuming, and clinics struggle to standardize practice across embryologists and sites. IVF clinics needed a comprehensive platform for automated embryo assessment, documentation, and decision support to standardize practices and improve outcomes.
What problem does Garbha.ai solve?
IVF clinics needed a comprehensive platform for automated embryo assessment, documentation, and decision support. AiSPRY engineered Garbha.ai with Rela Hospital around the structural limits of manual embryology practice.
Key challenges
- Grading subjectivity — visual morphological grading varies between embryologists, undermining consistency and benchmarking.
- Documentation burden — manual records of every assessment consume specialist time and invite gaps.
- Decision-support gap — embryo selection decisions lacked a standardized, data-backed second opinion.
- Clinical-grade data trust — a medical AI platform needs traceable data lineage and quality guarantees, not just model outputs.
How does Garbha.ai work?
AiSPRY built a complete AI platform for IVF clinics: a medical-imaging vision core that grades embryos automatically, a clinical workflow layer for documentation and decision support, and a data-governance backbone that makes every result traceable.
Embryology vision core
- YOLOv8 + Mask R-CNN imaging — detection and segmentation models assess embryo morphology from microscopy imagery
- ByteTrack tracking — embryos tracked consistently across frames and assessment sessions
- Roboflow data pipeline — curated, versioned training data behind every model iteration
- Automated grading — standardized grades at 90% consistency, replacing practitioner-dependent variation
Clinical platform and governance
- React clinical front-end — embryologists review gradings, documentation, and decision support in one workflow
- Comprehensive documentation — every assessment recorded automatically in PostgreSQL, cutting workflow time by 65%
- Data lineage & quality — OpenLineage + Marquez trace every result to its source; Great Expectations enforces data quality; Grafana monitors the platform
See Garbha.ai in action
A walkthrough of the Garbha.ai platform — embryo imagery through the YOLOv8 + Mask R-CNN grading core, automated documentation, and the decision-support view fertility specialists use.
Garbha.ai — standardized embryo grading and decision support
- Automated embryo grading — standardized morphological assessment at 90% consistency
- Built-in documentation — every assessment recorded automatically, 65% workflow time savings
- Decision support — data-backed second opinion for embryo selection
- Award-winning — winner of the 32nd HYSEA National Summit & Awards 2025
How does the platform handle clinical-grade requirements?
Embryology AI operates under clinical stakes that ordinary vision systems never face. AiSPRY engineered Garbha.ai around three constraints — assessment consistency, traceability, and the embryologist-in-the-loop.
Engineering constraints
- Assessment consistency — models grade against standardized criteria on curated, versioned training data, eliminating practitioner-to-practitioner variation
- End-to-end traceability — OpenLineage and Marquez record the lineage of every grading result; Great Expectations gates data quality at every pipeline stage
- Specialist-in-the-loop — Garbha.ai supports the embryologist's decision rather than replacing it; final selection always rests with the fertility specialist
What measurable results does Garbha.ai deliver?
The platform standardized embryo assessment at Rela Hospital and earned national recognition — moving both headline metrics decisively.
Headline outcomes
- 90% grading consistency — standardized AI assessment replaces practitioner-dependent grading variation
- 65% time savings — automated grading and documentation free embryologist time for clinical work
- HYSEA National Award 2025 — winner of the 32nd HYSEA National Summit & Awards, recognizing India's pioneering embryology AI
- Standardized lab practice — consistent assessment criteria and complete documentation across the IVF workflow
Garbha.ai — frequently asked questions
Below are the most common questions about how Garbha.ai works, what it grades, and how it fits into IVF clinical practice.