Syntra
Syntra, the operating name of Lightning Medical Corp., reviews medical claims against the underlying chart before submission. It reads codes and modifiers alongside visit notes, imaging, prior encounters and payer rules, flags missed charges and documentation that does not support a code, and cites the source so billing teams can check each suggestion. For eligible encounters it can apply approved edits and submit claims, and cases that need review go to a coder or biller first. Its coding review, compliance auditing and global period risk checks were trained with working medical coders.
A practice starts with a free retrospective audit, and the findings set its own guidelines. The 90 day rollout moves through controlled prebill review at day 30 and validation against the practice's data at day 60 before automation extends to eligible encounters. Integrations include Epic, Oracle Health Cerner, NextGen, athenahealth, eClinicalWorks, ModMed, Nextech, SIS and iMedicWare, and Pulse extracts structured data from scans and handwritten notes. Syntra sells into 22 specialties, from ophthalmology and dermatology to cardiology and oncology.
SightMD uses Syntra for prebill review, and an unnamed New York ophthalmology group with 37 locations saw net collections rise 1.33% and denials fall 26.5% in 2025 production use, about $1.58 million a year. Ayush Jain and Aniketh Kolla founded the Y Combinator backed company in 2024, and it sells annual subscriptions priced per physician, per chart or as a revenue share.
Capability Axes
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read
Syntra reads the claim alongside the chart and compares the documented services, diagnoses and modifiers with payer rules, flags an omitted procedure or shows the coder where an existing code lacks support, and acts on eligible clean encounters once a practice has approved the workflow. Each task depends on reading notes and claims together, including past visits and attachments, and the product is sold for exactly that review and submission work. Its audit reports, compliance reviews and billing copilot all run on findings from the same system.
A practice starts with a retrospective audit, turns confirmed findings into guidelines and runs controlled review before automation expands, and by day 90 the software works autonomously on eligible encounters. Clean eligible encounters go toward submission or posting while suggestions and exceptions go to billing staff, and teams validate against their own charts before adding encounter types and locations.
The homepage says the AI reasons on every chart like a human biller and captures missed coding opportunities across 100% of charts, coverage claims about which charts it reads. The eligibility cutoffs, an override path for a submitted claim and the conditions that call for a second reviewer are not stated.
The claims engine compares notes, imaging, codes, documentation and previous visits with the payer's billing rules, and its coding review was trained with medical coders. A separate extraction system turns scans and handwriting into structured fields with confidence scores and page locations, so the billing workflow can use document inputs and cite the original. The model behind the billing review is not named, nor how representative the coder training material is, and a practice has no shown way to see when a model version changes.
Pulse extracts fields from PDFs, scanned pages, images and handwriting sent by Syntra and returns structured data, bounding boxes and confidence scores for Syntra's billing models, with zero retention at that stage by Pulse's own account. That names one party in the document path and its role, apart from the model that checks codes or recommends billing changes. Claude and Gemini appear as approaches evaluated for ingestion, and no full list of model providers, or of the terms governing the downstream claim review, is given.
SightMD's Eric Rosenberg describes using Syntra to review encounters before billing. Syntra's own account of an anonymous 37 location ophthalmology group in New York reports 2025 production results of 1.33% higher net collections, 26.5% fewer denials and about $1.58 million in annualized revenue, roughly the collections of 1.5 more providers, without the claim count, how the comparison period was chosen or what other billing changes ran alongside.
An unnamed 12 provider group's audit estimates recoverable and future charges across 18 months of encounters, and the homepage claims $3 to $5 more per CPT code with autonomous review. None of it has independent validation.
The document pipeline sends selected billing documents to Pulse and gets back structured extraction with page locations and confidence scores, and Pulse and Syntra's chief executive both describe zero retention for those documents at that step. Syntra's own privacy policy covers electronic records, clinical notes and payment data held by the billing platform, a wider boundary than the extraction step, since the platform keeps chart information that Pulse discards. Syntra's full retention schedule, its training policy for clinical records and its handling of a reported AI safety incident are not stated.
The service privacy policy names Syntra a business associate to healthcare providers and covers its handling of electronic records, clinical documentation, billing data and other protected information, the FAQ says every account has a signed business associate agreement, and the eClinicalWorks integration page describes signing one before Syntra reaches a practice's data. The homepage calls the service HIPAA compliant. The signed form's terms are not public, including how its obligations reach outside extraction or model providers handling patient information.
Syntra states that its prebill review platform holds a SOC 2 Type II attestation and calls the service HIPAA compliant, and its FAQ describes encryption. Uploaded documents also pass through Pulse, which describes its own processing as encrypted and keeps no source data, a separate organization with its own assurance. Syntra's examination report, auditor and testing window are not public, and the scope across a customer's record interfaces, the extraction step and the claim review model is not drawn.
The scrubber checks clinical documentation against codes and payer rules for billing, detecting a missing modifier or a documented service left off a claim, then routing the chart for review or processing an eligible submission. It reads clinical material such as imaging records and past visits for that administrative purpose, and its checks against medical necessity requirements serve billing and denial risk. No FDA clearance is claimed, and the product is not sold to interpret findings for diagnosis or to direct treatment.
During rollout a practice compares findings with its own charts, sets local guidelines and validates performance against them before automation reaches more encounters. Coders get references to the supporting documentation, and a compliance team can review patterns of overcoding and undercoding across the practice, controls a customer can use before approving the system for another encounter type. No comparison of error or denial rates by patient group, payer, specialty or practice is offered, and no policy describes how such differences are tested and monitored once a practice is live.
A coder can follow a citation to the chart and decide whether a proposed billing change holds before the claim leaves the practice. The public contract covers intellectual property claims and caps ordinary contractual liability at 12 months of fees, and nothing in it reaches an incorrect billing suggestion or an erroneous claim submission. No output warranty or patient correction route is offered, and the case studies carry no error rate with the detail a practice would need to test it.
Prebill review integrates with Epic, Oracle Health Cerner, NextGen, athenahealth, eClinicalWorks, ModMed, Nextech, SIS, iMedicWare and Optivate, reading encounter notes and charges and submitting eligible claims through the customer's billing workflow, and a dedicated engineer builds a custom connection for unsupported systems. Some setups need workflow mapping during onboarding, and the eClinicalWorks setup signs an agreement before reaching practice data.
The full read and write operations for each EHR are not laid out, no production site is named with both directions verified, and claim submission can pass through a payer or clearinghouse outside the record.
Syntra is sold as a cloud service that connects to existing records, practice management systems and payer workflows. Pulse describes its document extraction API as supporting both cloud and local configurations, which speaks to the extraction component rather than to the option Syntra uses for a given customer.
The terms and privacy policy give no cloud region, no location for billing data and backups and no account of where the downstream models run, and nothing says a customer can run the billing platform on its own infrastructure.
The standard order form sets a one year subscription with monthly invoicing, priced by physician or by chart as a flat fee or a percentage of revenue, and the FAQ adds that practice size and chart volume move the price. A practice can start with a complimentary retrospective audit at no upfront cost, and a separate order may govern its commercial terms. No rate per physician or chart, revenue percentage or implementation price is shown, so the units help an administrator compare proposals while the paid subscription still needs a quote.
The scrubber is sold for outpatient specialty billing across 22 specialties, among them ophthalmology, optometry, dermatology, orthopedics, urology, cardiology, gastroenterology, oncology and multispecialty groups, and it checks the service, documentation and payer rules for each encounter, including laterality and global periods. Ophthalmology and dermatology deployments are described, and SightMD speaks to its use before submission. Measured figures come from one ophthalmology group, with no comparable results for the other specialties and no evidence that every record integration supports the same automated actions.
Alternatives to Syntra
Other products that do the same kind of work as Syntra, and what each one does.
Ember AI is an AI revenue integrity platform for specialty physician practices, surgery centers and health systems, sold on preventing denials upstream and recovering them downstream.
Inpatient medical coding and pre bill auditing software built on custom clinical AI and NLP, trained on millions of clinical records and tuned to each customer's patient population.
MDaudit sells a cloud platform for billing compliance and revenue integrity to health systems, physician networks, regional hospitals and ambulatory surgery centers, and says more than 100 US health systems with over $1 billion in net patient revenue use it.
Cair Health sells AI agents for billing teams that manage insurance claims from eligibility through payment. Chloe checks notes, coverage information and payer requirements before submission.
MedCare MSO offers Maximus, a practice management and billing application that connects scheduling, insurance eligibility, charges, claim submission, remittances and collections in one workspace.
AKASA, formerly Alpha Health, sells generative AI for the provider side of the revenue cycle to hospitals and health systems.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
Quote based
|
Per physician or per chart, flat fee or percentage of revenue | Signed BAA on every account | Not published; free retrospective chart and claim audit | Vendor Published |
The standard terms set a one year subscription with monthly invoicing, and practice size and chart volume affect a quote. No rate or revenue percentage is shown.