# Switching from InPractice AI, Dodon AI, and Claude: 4 stories > Four professionals switched to Medrecords AI in August 2026 from InPractice AI, Dodon AI, Claude, and human contractors. Their stories, de-identified. Canonical page: https://medrecords.ai/reports/why-experts-switch-to-ai-medical-record-review/ --- New — **Missing Records Detection:** flags every visit, provider, and date missing from the file. [See how →](https://medrecords.ai/product/missing-records-identification/) [Reports](https://medrecords.ai/reports/) › Why experts switch ## Why four experts switched from InPractice AI, Dodon AI, Claude, and human contractors In August 2026 four medical-legal professionals moved their record review to Medrecords AI: a nurse consulting team that left InPractice AI over duplicate-page billing, a consultant who left Dodon AI over misread VA records, a physician expert who could not defend Claude or Bastian GPT at deposition, and a forensic psychologist whose human contractors stopped delivering. By [Ahmed Jemaa](https://medrecords.ai/authors/ahmed-jemaa/) , Co-Founder & CEO of Medrecords AI · Published 2 Sep 2026 ### Four switches, one month This report tells four switching stories from August 2026. All four are real accounts, taken from recorded onboarding and support calls, with names, employers, and patient details removed. The customers are de-identified; the tools they left are not. One thing to hold onto while reading: everything said here about InPractice AI, Dodon AI, Claude, and Bastian GPT is the customer's reported experience, not a claim from the vendor's published record. For what each vendor publishes about itself, see the [InPractice AI review](https://medrecords.ai/reports/inpractice-ai-review/) , the [Dodon AI review](https://medrecords.ai/reports/dodonai-review/) , and the [alternatives pages](https://medrecords.ai/alternatives/) . The four had little in common. A nurse consulting practice focused on billing analysis. A legal nurse consultant working personal-injury files. An internist who testifies in medical-malpractice cases. A forensic psychologist evaluating conservatorship and conditional-release cases. They left different setups for different reasons, and the reasons rhyme. ### Story one: InPractice AI and the deduplication bill A consultant who runs operations for a busy legal nurse practice evaluated more than twenty chronology tools before choosing [InPractice AI](https://medrecords.ai/reports/inpractice-ai-review/) and prepaying roughly $4,000 in credits. The practice specializes in billing analysis for personal-injury cases, so its files run long: a recent case topped 2,000 pages and was, in his words, full of duplicates. He had real praise for the product, its side-by-side chronology view in particular. But in his account, the platform deduplicated only at a basic level and charged per page regardless, which meant the practice paid full rate for pages that were copies of pages it had already paid for. Worse, there was no way to check the work. “What we're using right now is a black box with that functionality. I have no way to see how you handled the duplication.” Operations consultant for a legal nurse practice, on InPractice AI, August 2026 What moved them was watching deduplication happen in the open: suspected duplicate pairs shown side by side so a reviewer confirms each match, page counts settled after duplicates are removed, and an export that keeps the table of contents, the chronology, and hyperlinked source documents in one file. Their next request, a chronology filtered down to just the billing records, became a filter instead of a feature request. ### Story two: Dodon AI and three notes on a page A legal nurse consultant working personal-injury cases hit a wall with [Dodon AI](https://medrecords.ai/reports/dodonai-review/) on VA records. VA files often put two or three separate notes on a single page, divided by nothing more than a horizontal line. In her account, the tool read each page as one unit: it merged unrelated notes into a single chronology entry under one provider, and it stamped entries with the admission date instead of the date the note was actually written. She corrected the dates by hand. She annotated the output PDF as she worked, and every time the chronology was regenerated her annotations were gone. When she raised it with the vendor, she says, the answer was that there was no fix. The cases she is building toward do not forgive that: a 9,000-page file now, and a 69,000-page nursing-home case behind it. At that scale, sub-page reading, note-level dates, and annotations that survive reprocessing stop being conveniences. They are the difference between a chronology she can stand behind and one she has to re-check line by line. ### Story three: Claude, Bastian GPT, and the deposition question An internist who reviews medical-malpractice cases at $650 an hour had been using Claude, the general-purpose AI chatbot, and found the analysis genuinely good. His words: Claude was amazing. The problem was not quality. A consumer chatbot signs no business associate agreement, and expert witnesses now get asked about their tools under oath: do you use AI, is it HIPAA protected, do you have a BAA. The wrong answers hand opposing counsel the cross-examination. So he moved to Bastian GPT, a HIPAA-wrapped chatbot service, and hit the other wall: the service's own support told him it could miss content in files beyond roughly 10,000 pages, he reported. His files are that big. One 4,000-page case arrived with no nursing notes at all, and he could not form an opinion until the gap was spotted and the missing records were requested. What he needed was completeness plus paperwork: [missing-records detection](https://medrecords.ai/product/missing-records-identification/) that flags what the file references but does not contain, a signed BAA, and [HIPAA-compliant processing](https://medrecords.ai/security/) he can describe plainly at deposition. ### Story four: when the contractors stopped answering A forensic psychologist with 25 years in practice evaluates conservatorship and conditional-release cases, on report templates the courts have seen from her for years. Her review pipeline was human: contract reviewers who worked up the file before she wrote. Then the contractors stopped being available, the summaries that did arrive were missing material, and there was no way to get the errors fixed. Her requirements were the opposite of generic. Her cases have no date of injury; the anchor is the date of the qualifying offense. Medication stability across the trailing year matters more than most of the medical file, and anything neurological is always relevant. She needed her two working report templates followed exactly, filters that speak her case types, and enough throughput for a full caseload. That is what her account became: her own templates rebuilt into the report studio, a forensic case type of her own, and drafts that follow her structure, which she edits and signs. The AI drafts; she decides. ### What the four switches share - › — **Verification over trust.** Every one of the four asked to see the chronology next to the source page before anything else. - › — **Billing that respects the file.** Counting pages after deduplication, with duplicates never charged, came up in every pricing conversation. Paying full rate for copies is the fastest way to lose a reviewer's trust. - › — **Coverage before opinion.** A chronology of an incomplete file is a liability. Flagging what the record references but does not contain figured in three of the four stories. - › — **Fit over features.** Custom templates, custom filters, custom case types: each switcher asked the tool to match how they already work, not the other way around. - › — **Paper that survives a deposition.** A signed BAA and HIPAA-compliant processing were entry conditions, not bonuses. ### How these stories were collected All four accounts come from recorded onboarding, demo, and support calls between these professionals and the Medrecords AI team in August 2026. Nobody was compensated and nobody was interviewed for marketing purposes; these were working sessions, condensed. Roles, specialties, file sizes, and the problems described are reported as stated. Names, employers, locations, and all patient information are removed or generalized. Where a customer described a competitor, that account is attributed to the customer, not presented as the vendor's published record; each vendor's own published claims are covered separately in the [vendor reviews](https://medrecords.ai/reports/) . See it on your own file The fastest way to test any of this is the way these four did: upload a real case and check the output against the source pages. [Test a file](https://medrecords.ai/test-a-file/?src=report-why-experts-switch-to-ai-medical-record-review) or [book a demo](https://medrecords.ai/demo/) . ### Questions about these stories #### Are these real Medrecords AI customers? Yes. All four accounts come from recorded onboarding, demo, and support calls held in August 2026. Roles, specialties, case sizes, and the problems described are reported as stated on those calls. Names, employers, locations, and all patient information are removed or generalized. #### Are the claims about InPractice AI, Dodon AI, Claude, and Bastian GPT verified? They are the customers' own accounts of their experience, reported as stated on recorded calls in August 2026, and they are attributed that way throughout. They are not assertions drawn from the vendors' published materials. For what each vendor publishes about its own product, see the vendor reviews at [/reports/](https://medrecords.ai/reports/) and the comparison tables at [/alternatives/](https://medrecords.ai/alternatives/) , which cite only published materials. #### Does Medrecords AI charge for duplicate pages? No. Self-service pricing starts at 10 cents per deduplicated page, and duplicate pages are removed and never charged. Suspected duplicates are shown side by side in the workspace so a reviewer can confirm each match instead of trusting a black box. #### What does de-identified mean in this report? The professional roles, specialties, file sizes, and quoted problems are real and unchanged. Everything that could identify a person or firm is removed: names, employers, cities, patient details, and any case fact specific enough to point to one matter.