Features
Everything below is one system with one login. Not a suite, not modules that sync overnight. The lead, the chart, the embryo and the invoice are the same record — which is why there is nothing to reconcile between them, and why the clerical hours a cycle costs today largely disappear.
Clinical
Most fertility EHRs are general-purpose charts with fertility fields bolted on. This one was built around the cycle, which is the unit a fertility clinic actually runs on.
The whole patient on one screen — no tab-switching to assemble a picture before you walk into a room.
Patient chart with trended labs and timeline.
Generated from a single date: birth control, stimulation, monitoring visits, trigger, retrieval, transfer. The patient sees the identical schedule in her portal — and the hour a coordinator spent building it, per cycle, is gone.
Generated calendar, Day 1 → 34, with protocol labels.
The screen a medical director opens at 7am. Sixty patients stimulating and you open only the ones that aren't tracking.
The patient side of the same record — so the schedule she reads is the schedule you built, character for character.
Embryology
Most EHRs stop at the chart. The embryology record — the one someone will ask you about in five years — usually lives in a spreadsheet nobody has permission to lose.
Retrieval outcomes through to transfer, per embryo.
Per-embryo record — grading, PGT, viability.
Down to the physical slot, so "where is that straw" is one click and not an afternoon.
Cryo drill-down: tank → canister → cane → straw.
Double witnessing where a person is structurally prevented from witnessing their own work — enforced by the system, not by a policy on a wall.
Funnel and money
The lead and the ledger sit at opposite ends of the same patient. In Solux they are the same record, which makes conversion reporting a query and month end an ordinary afternoon.
An eight-stage pipeline from first inquiry to treatment started, with every stage change timestamped.
Every name on the stale-lead list is a cycle you already paid to acquire and are about to lose.
Self-pay and insurance billing in one ledger, from day one — with the full US insurance path already built for when you need it.
AR aging and the billing ledger.
Compliance and operations
Register, calibration and preventive maintenance due dates, alarms with an acknowledgement trail, routine QC with supervisory review, proficiency testing.
Answers both recall questions — which cases used this lot, and which lots touched this case — without anyone opening a binder.
Credentials and competency assessments with expiry tracking, and the pre-inspection report generated rather than assembled.
The AI layer
A chat window is the easy half. The hard half is a model that can see the clinical record, is structurally prevented from seeing more than the person asking is allowed to, and leaves a trail a regulator would accept. That is a decision about the database, not a feature added later.
The assistant is the visible part. The rest of the AI in Solux runs without anyone opening a chat window — and all of it obeys the same constraint, because all of it reads the same permission-scoped database.
"Which patients have AMH below 1.0?" · "Who has retrievals this week?" · "Summarise this chart before I walk in." · "What's in AR past ninety days?" Forty pre-built questions across clinical, embryology, financial, operational, compliance and eligibility — and it answers well outside them too.
A full protocol from one date, with a pre-flight check that refuses to generate rather than produce a schedule built on missing scheduling rules or provider availability.
Outside records and lab reports parsed on upload, fields mapped into the chart, and anything below confidence routed to a human instead of written quietly.
Clinical thresholds, equipment and tank alarms, failed QC, credential expiries, stock-outs and aged receivables — surfaced into one inbox with an owner, a routing rule and a resolution trail.
Five dimensions per patient — clinical, laboratory, consents, financial, operational — so the Command Center can tell a director which four of sixty patients need him today.
Input features, predicted value, confidence and model version are stored on every prediction — and the actual outcome is written back beside it. You can audit whether the model was right, by type and over time.
Population-level findings arrive with a confidence level and a status, and stay in review until a named clinician signs them off. Nothing acts on a patient because a model suggested it.
English and Spanish across the app, the portal and every AI surface — with machine-critical fields (event types, medication names, field mappings) deliberately excluded from translation so a locale switch can never alter clinical meaning.
Structured daily summaries generated off the live record, and AI-assisted intake conversations that arrive as data rather than as a transcript someone has to re-key.
Name, persona, tone and clinic knowledge are configured during onboarding. It introduces itself as your clinic's assistant, not ours.
And to be exact about what that is: prompt and knowledge configuration, not a model trained on your patients.
Integrations
These are not roadmap items and they are not quoted separately. They exist, they are running, and connecting yours is a configuration step during onboarding.
Chain-of-custody events written straight into the specimen record.
Embryo development data and AI viability scores against the individual embryo.
Ultrasound studies over a site-to-site tunnel, attached to the chart.
Results in and orders out over a standard HL7 v2 interface.
Card and ACH against the invoice that generated the charge.
Transactional patient messaging in English or Spanish.
Because parts of this industry still run on both.
Full English and Spanish across the app, the portal and the assistant.
Implementation
Forty configuration items across identity, scheduling, clinical, embryology, billing, CRM, people, payments, lab QC, inventory, workforce, revenue cycle and integrations. Each one shows its status, whether it's a blocker, why it matters in one plain sentence, and exactly which screen fixes it.
When every blocker is green, you go live. That's the whole project — visible to you, in real time, from the first week.
Your cycle volume, your systems, your handoffs, your numbers. Forty-five minutes, before anything is configured.
Your own database, storage and domain, deployed from a versioned template. Days, not months.
Prices, protocols, consents, staff and scheduling rules — run against the 40-item checklist with us in the room.
Structured import with a review step before anything becomes a clinical record. Training by role, not one long generic session.
Blockers green, patients on it, and a named person who answers the phone afterwards.
Next step
The best first question isn't whether you like the software. It's which part of your current week you'd want running on this first.