Optexity

We're launching APIs for desktop-based EHRs today - our most requested feature! Desktop based EHRs such as Epic, Cerner, MEDITECH, dominate the healthcare landscape. API access is often limited or non-existent due to custom builds. Hence we are releasing desktop automation today, to build APIs for these systems. Now you can automate any system involving RDP, Citrix, or Windows machines. Our AI agents learn to perform the workflow just from your screen recordings and use native computer vision to navigate the screen. Pre-training these agents on your workflows makes them very fast and highly accurate so that you can run them tens of thousands of times a day in a cost effective manner. Here is a demo of downloading 3800 faxes a day from Raintree using our desktop automation - which our customers are already using! If you're building on Epic, Cerner, MEDITECH, or any other desktop-based EHR, you can get started at dashboard.optexity.com or DM us for onboarding help.

RPM/CCM billing has a timing problem that most RCM teams don't talk about. Many of these codes can only be billed once per patient per month. If your team is processing claims in end-of-month batches, and another provider submits first, you've lost that revenue permanently. Not delayed. Gone. On top of that, someone has to manually reconcile device data, accumulated care time, eligibility, and documentation across multiple screens in @eClinicalWorks . Every single month. For every single patient. Here is a real workflow running in production showing how @Optexity automates this entire workflow. You send patient data, and Optexity logs into eCW, monitors billable thresholds as they're met throughout the month, compiles the claim details (demographics, ICD-10 diagnoses, CPT codes), and submits the claim the moment it's ready. No batch cycle. No manual reconciliation. No timing risk. It works directly inside eCW without requiring any API access, custom interfaces, or workflow changes on your end. The result is continuous billing instead of a stressful month-end scramble. Clean-claim rates go up, revenue leakage goes down, and your staff stops spending their day navigating legacy portals. If your RCM team is still processing RPM/CCM claims manually, you can get started for free at dashboard.optexity.com.

Every time an insurance payment comes in, someone on your billing team has to open the EOB, figure out which patient it belongs to, find the right claim in @drchrono by EverHealth, and then post the payment field by field. Payer name. Check number. Paid amount. Contractual adjustment. Patient responsibility. Denial codes. Across every single CPT line. At 30 claims a day, that's hours of pure data entry. And if someone makes a mistake, wrong adjustment amount, missed write-off, payment posted to the wrong service line, you're dealing with claim errors that take even longer to unwind. This is what we built @Optexity for. You send us an EOB, and our automation logs into DrChrono, finds the right patient and claim, and posts every payment detail line by line. Adjustments, write-offs, and patient responsibility applied exactly where they need to go. The ledger updates itself. Your billing team doesn't touch the portal. If your team is still doing this by hand, you can try Optexity today for free at dashboard.optexity.com.

Before a claim gets filed, someone has to verify the patient's coverage on insurance portals like Ambetter. Is the insurance active? What's the deductible status? What's the copay? What's the OOP max? Ambetter doesn't just surface this on a single screen. The real coverage detail lives inside a PDF, a benefits schedule buried in the portal that someone has to open, read through, and manually pull numbers from. At busy practices, this is happening hundreds of times a day. Staff logging in, searching the patient, downloading the PDF, hunting for the right line items, copying them somewhere useful. Every single patient. Every single visit. We built a demo showing how one of our customers is automating this entire workflow with Optexity. You send a patient, and @Optexity logs into the Ambetter portal, looks them up, verifies eligibility, opens the benefits PDF, and extracts the structured data you actually need. Deductibles, copays, coinsurance, OOP max, all returned as clean, usable output. And this is where it goes beyond just browser automation. Optexity natively handles document extraction, intelligent PDF analysis, and OCR as part of the same workflow. So when that benefits schedule PDF opens inside the portal, we don't just download it and hand it off to another tool. We read it, parse it, and pull out the exact fields you need, all within the same automation run. Your team doesn't have to touch a separate pipeline, stitch tools together, hunt through PDFs, or copy numbers into spreadsheets anywhere. You send a patient, and the whole thing runs end to end, automatically. We're in production serving companies across the US healthcare system right now. If your team is still doing this by hand, you can start using Optexity today. Open source. Free to get started. Reach out if you want help getting set up.

Before a doctor can submit a prior authorization, they need to check if the patient's insurance is even active. They need to know the copay, the coinsurance, the deductible status, all of it. Because patients need to be charged correctly and they deserve to know what they're paying upfront. This isn't optional. It's a required step in the workflow. Most clinics pull this data from clearinghouses like @Availity . Real time eligibility verification, benefits information, coverage details. It's all there. But there's a problem. @Availity doesn't have an API. Which means staff have to log into the portal manually. Search for each patient. Pull the eligibility data. Copy it into their system. Then move on to the next patient and do it all over again. When you're running 30, 40, 50 patients a day, this adds up fast. One clinic told us their front desk was spending close to 2 hours every day just on eligibility checks. Not answering calls, not scheduling appointments, not helping patients. Just logging into Availity and copying information. That's almost 10 hours a week. Per person. On something that should take seconds. This demo shows what we at @Optexity built to fix it. You send patient details to Optexity, and it handles the entire @Availity lookup automatically. What comes back is structured data you can actually use. Member status, total deductible, remaining deductible, PCP copay, specialist copay. All formatted and ready to plug into whatever system you're running. @Availity doesn't offer an API. But now you have one anyway. The data flows out automatically, and your team never has to touch the portal again. We're in production now serving companies across the US Healthcare System. If your team is still doing eligibility checks manually, you can start using Optexity today. Open source. Free to use. Reach out if you want help getting set up.

It’s been a while since I recorded a proper video like this, so I am really excited to share this!!!! New role. New space. New problem to solve. I recently joined @Optexity as a Founding DevRel Engineer, and I finally got around to recording a walkthrough of what we’re building and why I find it so exciting. In this video, I talk about a very real gap in healthcare automation. AI scribes today can listen to doctor patient conversations and generate structured notes beautifully. But after that, someone still has to manually enter those notes into the EHR system. That part is slow, repetitive, and prone to human error. Most legacy EHR systems were never designed with APIs in mind. They were built for humans to click, type, and submit forms. And yet, these systems are still deeply embedded across hospitals and clinics. So how do you automate something that does not expose an API? That is exactly what we are solving at Optexity In this demo, I show how AI scribe notes are directly uploaded into the @practicesuite EHR system using Optexity, without a human touching the keyboard. No copy paste. No manual entry. Just automation working with the existing portal. If you are building in healthcare, working with legacy systems, or thinking about agentic automation in the real world, I would genuinely love for you to check this out. Also, it feels really nice to be back to recording and sharing again. Let me know what you think of the demo.

We're already in production serving over 200k requests a month. If you're dealing with this in your practice, you can start using Optexity today. We're open-source and free to use. Reach out if you would like a personal onboarding session.

This demo shows what we built - scribe output flowing into @practicesuite with zero manual steps. Visit ends, note uploads itself, doctor moves on. No extensions, no "did I remember?", no extra clicks. Just the workflow completing itself in cloud automatically.

US healthcare is used to AI scribes by now. We've been watching this space, and noticed something interesting - everyone's focused on making transcription better, more accurate, faster. All good stuff. But nobody's fixing what happens after. 🧵

@dharmesh @DylSell happy to hear your your thoughts on copilot for hubspot :)

Super excited to share what we’ve been working on – Optexity, a platform for training browser automation agents that work reliably on any website. With our platform you can train agents by showing human demonstration of your workflows.

With the current repository you can create human demonstrations from which models can learn to use that software. Or simply try vanilla out of the box gemini to complete tasks on a website. Excited to see how devs use it to make their agents reliable.

We fine-tuned Llama 3-8B (10x smaller and cheaper), outperforming GPT-4o & Gemini 2.0 on complex @HubSpot workflows. We will be adding more functionality to train agents from RL, docs & YT tutorials to reliably operate software. Demos and source code is at optexity.com