Quick answer: You don't rip out RightFax to automate inbound fax routing in NextGen — you layer an AI document classifier on top of it that reads each fax's actual content (not just the cover sheet or fax number), routes by document type and patient match, and writes back to NextGen through the API or interface engine. Your existing RightFax deployment keeps handling transmission, your barcode and DID routing rules stay live, your staff keeps the same NextGen work queue, and the AI classifier becomes the layer that decides where each fax actually belongs based on what's on the page. Implementation runs 4–8 weeks for most NextGen Enterprise practices.
Why your RightFax setup hits its limits even with barcode and DID routing
Most NextGen Enterprise practices that ask about fax triage automation are already running a credible RightFax setup. Inbound fax numbers route by Direct Inward Dialing (DID) to specific work queues. Barcode routing pulls cover-sheet barcodes off departmental forms. Smart Fax Distribution sends faxes to members of a group based on availability. The fax server is doing its job.
What it can't do is read the document. Once the fax lands in the right departmental queue, a staff member still has to open it, identify the patient, classify what kind of document it is, decide which template-folder in NextGen it belongs to, and route any follow-up tasks. The 8–15 minutes per complex fax doesn't change because the fax server isn't the bottleneck — the document interpretation is.
The math gets uncomfortable at multi-provider NextGen Enterprise practices. Industry data puts the share of clinical documents arriving by fax at 35–45% of inbound volume, with the healthcare industry exchanging over 9 billion fax pages annually. For a 25-provider multi-specialty group, that translates to 8,000–15,000 inbound faxes a month after the fax server has done its routing. Each one still needs a human to interpret it.
The instinct to fix this by replacing RightFax is the wrong instinct. RightFax handles the transmission layer well. The work that costs you hours is downstream of RightFax, not inside it. Layering an AI classifier on top of the existing fax server is the cleaner architectural move — fewer integrations to manage, no number changes, no retraining your team on a new fax UI.
The integration architecture: RightFax → AI classifier → NextGen API
Modern AI fax triage for NextGen Enterprise practices runs as a three-stage pipeline that sits on top of RightFax rather than replacing it.
Stage 1 — RightFax keeps owning transmission. Inbound faxes arrive on your existing fax number through your existing DID routing, barcode rules, and Smart Fax Distribution configuration. Nothing changes for the referring practice — they send to the same number, with the same cover-sheet conventions, and receive the same delivery confirmations. RightFax does what it has always done: receive the fax, apply your routing rules, and drop the PDF into the appropriate departmental queue.
Stage 2 — AI classifier reads the content. The triage layer monitors RightFax's outputs (typically via a folder watch, an HL7 feed, or RightFax's REST API) and processes each PDF as it arrives. A healthcare-trained classifier identifies the document type — referral, lab result, prior auth response, refill request, records release. An extraction model pulls structured fields: patient name, DOB, MRN, referring provider, diagnosis codes, requested service. A patient matcher identifies which NextGen chart the document belongs to, with a confidence score.
Stage 3 — Structured write-back to NextGen. High-confidence matches write into NextGen through the API for cloud-deployed practices or through HL7 v2 messaging plus an interface engine for on-prem NextGen Enterprise. The document lands in the correct patient chart, in the correct template-folder, with the document-type tag the auth or clinical team expects. Follow-up tasks route to the right NextGen user queue. Low-confidence cases queue for human review with the AI's best guess pre-populated.
The end-to-end latency from fax arrival on RightFax to filed-in-NextGen-with-task-routed runs 60–90 seconds at production volume. Your IT team doesn't have to modify your fax server. Your auth team doesn't have to learn a new UI — they still work out of NextGen, but the work shifts from manual classification to exception review.
How the AI handles your three existing routing patterns
NextGen Enterprise practices on RightFax typically run a mix of three inbound routing patterns. Good triage software respects each one and adds intelligent classification on top, rather than fighting your existing setup.
Multi-location DID routing. Your central RightFax instance answers multiple fax numbers — one per location, sometimes one per department within a location. Routing rules send each inbound fax to a location-specific or department-specific work queue. The AI classifier reads each fax's content and decides what to do with it after RightFax has identified the destination location. A cardiology referral that arrived on the cardiology DID at the Westside location still routes to Westside cardiology — the AI just identifies the document as a referral, pulls the patient demographics, matches to the right chart, and creates the scheduling task. The location-routing decision RightFax already made stays intact.
Smart Fax Distribution rules. Your group routes faxes within a department by member availability — a fax addressed to "Cardiology" lands in the queue of whichever cardiology MA has the lightest workload. The AI layer doesn't interfere with that load-balancing decision. It processes the document after distribution and writes the structured filing back to NextGen, so the assigned MA opens NextGen and finds a pre-classified document with the chart already attached, instead of starting from a raw PDF.
Barcode routing on departmental forms. Your practice prints cover sheets with barcodes that route specific document types — incoming records requests, prior auth response packets, lab corrections. RightFax reads the barcode, applies the routing rule, and drops the fax in the right queue. The AI layer sees the barcode, recognizes the document type from the barcode metadata, and uses that as a high-confidence input to its own classification. Where the barcode is missing or unreadable, the AI's content-based classifier takes over.
The architectural principle: existing routing rules win when they're present, AI fills the gaps when they aren't, and the two layers always agree on the same destination because the AI reads RightFax's routing decision before making its own.
What to instrument and watch in the first 30 days
Most NextGen Enterprise practices that adopt AI fax triage underestimate how much measurement matters in the first 30 days. Three metrics tell you whether the system is working — or whether your team is silently working around it.
First-pass classification accuracy. Pull a daily report of how many inbound faxes were classified at high confidence versus low confidence. Industry-strong systems hit 96–99% high-confidence classification at full production once tuned. If your numbers run below 90% past day 14, the classifier hasn't tuned to your specific document mix yet — push the vendor for additional training on your edge cases.
Time-to-chart latency. Measure median elapsed time from RightFax delivery to NextGen filed-with-task. Steady state should land at 60–90 seconds. If it stretches past two minutes, the bottleneck is usually in the NextGen write-back layer rather than the AI itself — check whether your interface engine is processing the load and whether NextGen's API rate limits are bottlenecking.
Manual touch rate. Track the share of inbound faxes that any human touched — opened, reclassified, manually filed, or routed. Pre-automation this was 100%. By week 4 it should be 15–25%; by week 10 it should be 5–15%. If it stays high past 30 days, your team is silently working around the AI rather than trusting it — usually because the exception queue UI isn't designed for fast review, or because change management didn't land.
A 2024 industry survey found that 52% of faxes require staff intervention in non-automated environments. If your number doesn't fall meaningfully below that baseline within the first 30 days, something in the pipeline isn't working as intended.
The change-management piece nobody warns you about
The technical work of layering AI fax triage on top of RightFax is straightforward. The harder problem is getting your front desk and auth team to stop sorting faxes the old way.
Your team has built muscle memory around the manual workflow. When a fax lands in a NextGen queue, the auth coordinator opens it, reads it, finds the patient, classifies the document, and files it. That sequence runs without conscious thought. After AI triage goes live, the same coordinator opens NextGen and finds the document already classified, the patient already matched, the follow-up task already routed. The natural human response is to re-verify — to open the PDF, check the AI's work, and basically redo the classification mentally before trusting the system.
If you let that pattern persist, you'll capture maybe 30% of the labor savings the technology is capable of delivering. The team is still doing the cognitive work; the AI just did it twice.
The fix is a structured change-management plan during the first 60 days:
- Run shadow mode for 7–10 days. The AI processes faxes; your team continues to do their normal manual work in parallel as a control. Daily reports compare AI decisions to human decisions on the same documents, building team confidence in where the AI is reliable.
- Set explicit trust thresholds. Documents the AI classifies at >95% confidence go directly to filed-and-tasked status; the team only reviews the exception queue. Documents at 85–95% confidence get a quick visual check (10–15 seconds) before being approved. Below 85% gets full review.
- Track time-to-chart per team member. When you can see who's still opening every PDF versus who's working from the AI's filing, you can have a targeted conversation rather than a general one.
- Tie the change to a benefit visible to the team. The recovered hours don't disappear — they redeploy to prior auth follow-up, denial appeals, or referring-provider outreach. Make that redeployment visible. Teams trust automation more when they see the time savings going to work they actually want to do, not to headcount cuts.
Honey Health's Fax Triage agent is built to layer cleanly on top of RightFax and NextGen Fax Manager environments — content-based classification, multi-signal patient matching, structured write-back through the NextGen API or interface engine, and a confidence-scored review queue designed for fast triage rather than full reclassification. The agent doesn't ask your team to learn a new tool; the work shifts to exception review inside NextGen.
How this differs from ripping out RightFax for a cloud-native fax service
Some vendors will pitch you on replacing RightFax with their cloud-native fax service plus AI on top. That's a credible architecture for greenfield practices, but it's usually the wrong move for a NextGen Enterprise practice with an established RightFax setup.
Three reasons to keep RightFax and add the AI layer rather than replace:
- Your fax numbers are sticky. Every referring practice has your fax numbers stored in their EHR's address book. Notifying them, getting them to update, and confirming the change stuck takes 3–6 months at meaningful scale. During that window, some referrals route to dead numbers. The risk-adjusted cost of a fax migration is much higher than the subscription savings.
- RightFax features you depend on don't always replicate cleanly. Smart Fax Distribution, barcode routing, encrypted internal fax queues, and on-prem audit logging are mature in RightFax. Cloud-native fax services replicate some of these but rarely all, and the migration usually exposes a gap your team has been depending on for years.
- Procurement and security review for a new fax vendor is expensive. RightFax is already in your BAA portfolio with HIPAA, HITRUST, and SOC 2 documentation completed. A new vendor restarts that work. For most NextGen Enterprise practices, the AI triage layer is the simpler procurement story because it slots into an existing vendor relationship rather than starting a new one.
The honest framing: replace RightFax only when your fax server is genuinely failing your operational needs. If it's working and the bottleneck is downstream document handling, layer the AI on top.
Frequently asked questions
Will adding AI fax triage on top of RightFax require any RightFax configuration changes?
Usually minimal. Most triage vendors integrate through one of three RightFax surfaces — a watched output folder, RightFax's REST API, or an HL7 feed — and don't require changes to your existing DID routing, barcode rules, or Smart Fax Distribution configuration. The integration is additive, not destructive. Your fax server administrator typically spends 4–6 hours during initial integration and minimal time afterward.
How does the AI handle barcode-routed faxes where the barcode metadata already tells RightFax the document type?
The AI uses the barcode metadata as a high-confidence input to its own classification rather than ignoring it or duplicating the work. A fax that RightFax routed as "records request" via barcode lands in the AI pipeline already tagged; the AI extracts the patient demographics and clinical data, matches to the NextGen chart, and files with the records-request tag. Where the barcode is unreadable or missing, the AI's content-based classifier takes over.
What happens if the AI's classification disagrees with our existing RightFax routing rule?
Strong triage systems surface the disagreement to a human reviewer rather than overriding silently. A fax that RightFax routed as a referral but the AI classified as a records request gets flagged in the exception queue with both candidate classifications visible. The reviewer makes the final call. Over time, these flags identify rule-set drift — places where your RightFax routing logic and the AI's content-based reading consistently disagree — which usually points to barcode rules or DID assignments that need updating.
How long does implementation take for a NextGen Enterprise practice with an established RightFax setup?
Most NextGen Enterprise practices reach go-live in 6–10 weeks. The technical integration with RightFax usually finishes in 1–2 weeks; the NextGen Enterprise write-back through HL7 messaging and an interface engine adds 3–5 weeks because Enterprise integration is heavier than NextGen Office; AI tuning to your specific document mix and patient database adds another 1–2 weeks. The change-management work runs in parallel.
Will adopting AI fax triage on top of RightFax require us to expand our HIPAA BAA portfolio?
Yes, you'll need a Business Associate Agreement with the triage vendor that covers PHI handling, audit logging, and breach notification. Strong vendors layer on HITRUST CSF certification and SOC 2 Type II audits, and their BAA template should be straightforward to review. Your existing RightFax BAA stays in place; the new vendor's BAA is additive. Procurement review typically runs 2–4 weeks at most NextGen Enterprise practices that already have a mature vendor-risk-management process.
