AVA reads scanned requisitions and insurance documents, extracts patient and order data with AI, and routes every record through a confidence-scored human review, so your lab syncs clean data faster.
It is 8:02 on a Monday.
Sixty faxes came in over the weekend. Each one needs a person to read it, retype it and hope nothing was misread.
What if the typing was already done?
Read it, retype it, check it, fix it. Each slip below is a pain labs live with, and what AVA does about it.
Every faxed requisition is read by a person and retyped field by field. Your best people spend the day as typists.
AVA extracts patient, insurance, provider, panels and ICD-10 codes from the requisition and hands them over ready to check, not to type.
The requisition, insurance card, ID, face sheet and clinical notes arrive separately and get matched by hand.
AVA merges every document for a patient into a single record and keeps the source of each value.
A transposed digit or a misspelled name travels from intake to billing, and comes back weeks later as a denial.
Every extraction carries confidence scores. Handwritten dates, fax noise and conflicting data are marked in red before anyone approves.
Black-box tools push data straight in. When they are wrong, nobody knows until it hurts.
Nothing becomes a record until a reviewer approves it, with the original document right beside the data.
Generic tools guess test codes. Unknown panels slip into orders as free text nobody mapped.
Panels, analytes, containers and collection methods come from your own master data. Unknown codes are flagged, never invented.
An order exists but the patient does not, or the insurance is missing. Someone cleans up for days.
Approval creates the patient, insurance, medications, allergies, order, panels, analytes, ICD-10 codes and samples in one step.
Which faxes are pending, which failed, which need a second look? You find out when a client calls.
Live queues, a transaction log and analytics show what is pending, processing, verified, failed or in review.
Different labs, clients and sites each need their own rules, storage and branding.
A multi-tenant hierarchy gives each lab its own configuration, intake folders and AI settings.
8manual steps retired. One verified intake.
Try it. Pick a field to see what AVA extracted, then move the threshold to decide how sure the AI must be before a record skips the review-first queue. Demo data only.
If a field cannot be read, it is left empty and confidence is lowered. AVA does not guess.
First name, last name, date of birth, gender and at least one test must be present, or the record goes to a reviewer.
Choose how sure the AI must be, per portal. Everything below it is routed to a person.
An illustration of how a single faxed requisition moves along the line. The struck-through text is how it usually goes.
It lands in a fax inbox and waits for someone to open it.
With AVAIt is picked up from your secure intake folder, checked for duplicates and queued.
Someone clips the card and ID to the requisition.
With AVARequisition, insurance card, ID and face sheet are grouped as one patient.
A person types each field, squinting at faded fax text.
With AVAPatient, insurance, provider, panels and ICD-10 codes are extracted, each with a confidence score.
The reviewer re-reads the whole form hunting for mistakes.
With AVAThe original sits on the left, the extracted record on the right. Uncertain fields are highlighted.
Records are keyed in several screens, and sometimes missed.
With AVAPatient, insurance, order, panels, samples and codes are created together, with an audit trail.
Someone emails a spreadsheet to the lab system.
With AVAThe record moves from draft to QA review to ready, and your lab system acknowledges the sync.
Someone counts faxes in a spreadsheet.
With AVAAnalytics show volume, approvals, confidence and processing time.
A focused workspace for checking, correcting and approving, and a view of how the automation is doing.
Illustrated screens use demo data.
Six stages, every capability. Open any item to see how it works.
Files arrive securely and never get lost or doubled.
Per-portal SFTP folders and direct upload, with connection testing, remote browsing, preview, rename and download.
Files are hashed for integrity and duplicate checks, with pending, processing, processed and failed status tracking.
Pull files into cloud storage with live progress, cancel support and file counts. The cloud file browser stays available even when SFTP is switched off.
AI does the typing. Confidence scores tell you where to look.
Requisitions, insurance cards, IDs, face sheets and clinical notes are merged with source attribution.
Overall, patient and order confidence plus granular flags. Anything under your threshold goes to review, and missing mandatory fields always do.
No invented values. Illegible fields are left empty and confidence is lowered. Unknown codes are flagged as free text. Temporary AI copies of documents are cleaned up after processing.
Set routing, fallback, confidence threshold, timeout and retries per portal, with performance and cost tracking and a global off switch.
A focused workspace for checking, correcting and approving.
Pending, reviewing and approved tabs, portal and date filters, column search, and live counts including low confidence.
Original document on the left, fully editable extracted data on the right. Save corrections without creating records.
Demographics, guardian, insurance, medications and allergies with autocomplete, provider lookup, panels, add-on analytes, ICD-10 codes and specimens by container.
Approval creates all linked records in one step. Rejections go to manual review with a reason, and every correction and approval is logged.
Panels, analytes, containers, codes and providers from your own catalog.
Grouped test configurations, portal-specific mappings and add-on analytes, searchable and bulk capable.
Collection methods, container types, collection details and auto-generated sample IDs, with panels mapped per sample.
An ICD-10 code library, plus medication and allergy libraries used throughout orders and review.
Provider directory with NPI search and NPI and PECOS verification, linked to sites.
Clear statuses and APIs so your lab system gets verified data.
Records move from draft to QA review to ready to syncing to synced. Synced orders lock to protect data integrity.
Status endpoints let your system mark orders and patients as syncing or synced, with an external reference and notes.
Token-secured APIs for automation files and requests, direct patient and order creation, and lookups for panels, analytes, providers and codes.
Multi-tenant, controlled, audited and measurable.
Portal to client to site to patient to order to sample, with per-portal configuration, branding and storage.
Roles with granular permissions, session and optional two-factor controls, audit logs and login activity.
Branding, timezone, mail, storage, automation and security settings are managed in a configuration screen, not in files.
Eleven or more views: requests over time, status and confidence distribution, processing time trends, daily approvals, hourly volume, retries and more.
AVA sits between the people who send requisitions and the system that runs your lab. We connect it to the systems around you. Scope depends on those systems, so we start with a discovery call.
Providers keep sending requisitions the way they always have. AVA picks them up from a secure folder or direct upload.
Provider and address checks happen during review, not after a rejection.
Your LIS, LIMS, EHR or EMR receives verified records and reports back when they are synced.
Requisitions are full of protected health information. AVA is designed to support HIPAA requirements at every step, from intake to approval.
Four default roles and 18 granular permissions, with a screen to assign them.
Audit logs on orders, corrections and approvals, plus login activity.
Configurable session lifetime and optional two-factor sign-in.
Hash-checked files in secure storage, served through protected routes.
Temporary AI copies of documents are removed after processing, with a cleanup for anything left behind.
Synced orders lock automatically, and records are created in a single step so nothing is left half-built.
We set AVA up around your documents and test it with your team before it touches live work. Here is who does what.
We learn how requisitions reach you today, which documents you receive and where errors creep in.
Portals, clients, sites, intake folders, catalog mappings, roles and AI settings are set up around your lab.
Real requisitions go through review with your team, and we tune confidence thresholds with you.
Your team is trained, and the engineers who built your setup stay close as you go live and after.
Nothing to install. Your team signs in from any modern browser.
Role-based access, audit logging and controlled sessions.
Branding, storage, intake and AI settings for each lab.
Uncertain data always goes to a reviewer before it becomes a record.
API access for the systems you already run.
Our senior team helps you configure, test and go live.
We are not just selling a service. We are building trust and making lab operations faster. Real people answer, quickly, and a dedicated Project Manager is always there for anything you need.
No robots. Real humans answer you.
Your team and ours in one channel.
Reach support and technical teams directly.
Pick up the phone when talking is faster.
AVA (Accession Via Automation) is Jeevaysoft’s intake platform for laboratories. It reads faxed, scanned and uploaded requisitions and insurance documents with AI, routes every record through a human review workspace, and prepares clean patient and order records for your lab system.
No. Every extraction carries confidence scores, uncertain fields are highlighted, and nothing becomes a patient or order record until a reviewer approves it.
Requisitions, insurance cards, government IDs, face sheets and clinical notes. Documents for one patient are bundled into a single record, with the source of each value kept.
Through a secure SFTP folder for each portal, or by direct upload. Files are checked for duplicates and tracked from pending to processed.
Low-confidence fields are flagged, illegible values are left empty rather than guessed, and records with missing mandatory fields always go to review.
AVA sits in front of your lab system and hands over verified records. Your system can report sync status back through our API. Scope depends on the systems involved, so we start with a discovery call.
Yes. AVA uses your own catalog of panels, analytes, containers and collection methods. Unknown codes are flagged for review instead of being invented.
Yes. AVA is multi-tenant, with portals, clients and sites, and separate configuration, storage and AI settings for each portal.
AVA is designed to support HIPAA requirements, with role-based access, audit logs, controlled sessions, optional two-factor sign-in and file integrity checks. Jeevaysoft is ISO/IEC 27001:2022 certified and our team is HIPAA trained. Compliance documentation is available on request.
Live chat with real people, a dedicated MS Teams group, email and phone calls. Our support and technical teams respond quickly, and a dedicated Project Manager is available for any kind of assistance.
Yes. Book a demo and we will walk through the review workflow with your kind of requisition, and discuss anything custom you need.
Send us a few of your real requisitions. We will show you what AVA extracts, where it is unsure, and how review works for your team.
