SmartDepo
SmartDepo turns deposition transcripts into page-line summaries with hyperlinked citations back to the source testimony, intended to be copied straight into motions and briefs. Every transcript produces four documents: a page-line summary organised by topic and running a seventh to a tenth of the original length, a key admissions analysis of one to five pages, a short deposition abstract, and a deposition memo organised thematically around the legal and factual issues. A newer platform layer adds chat across every deposition in a case, automated memos, a contradiction finder and theme extraction. The company describes its citation technology as patent-pending and says every summary is reviewed by legal professionals before delivery. Pricing is published in full and can be bought two ways: a 99 dollar per user monthly subscription that reduces summaries to a flat 25 dollars each, or standalone summaries from 85 dollars falling to 50 at volume, with a seven-day free trial that needs no credit card. SmartDepo is now part of Rev, and the agreement a buyer accepts on the SmartDepo site is Rev.com, Inc.'s terms of service, which govern Rev's transcription, captioning and translation services and do not mention SmartDepo by name. Those terms state that customer content may be used for continuous training of Rev's own speech recognition and other AI models, though not for generative AI training, and commit Rev to maintaining professional errors and omissions and cyber liability insurance.
Capability grades
All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
The models are the mechanism the buyer pays for. What is sold is an AI deposition summary produced in minutes from a transcript, described by the vendor as resting on patent-pending technology, and the newer platform layer is entirely model work: chat across every deposition in a case, a contradiction finder, theme extraction and automated memos. Human quality review sits on top of the model output as a check rather than as the production method, which is the distinction that keeps this at the top of the band. Remove the models and what remains is the manual page-line summarisation the product exists to replace. Checked 4 September 2026.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Grounding is real and documented: summaries carry hyperlinked page-line citations to the source dialogue, the vendor publishes a side-by-side comparison of its own precise citations against competitors' broad topic ranges, and every summary is stated to be reviewed by legal professionals before delivery. Those are architectural and procedural controls rather than a claim alone. What sits against them is the strongest marketing-versus-agreement conflict on this record, and the agreement governs. The site advertises 100 per cent guaranteed accurate page-line citations and patent-pending technology that can guarantee 100 per cent accuracy. The governing terms at the same domain state that the Services may contain errors, disclaim any warranty as to the quality, accuracy, currency or completeness of the platform or any results obtained through it, and make the customer solely responsible for verifying the accuracy and completeness of all work product before acting on it. A guarantee that the agreement expressly disclaims is not a measured accuracy figure, no test set is described, and the failure modes the agreement does name are audio quality, background noise, cross-talk and speaker accent, which belong to transcription rather than to summarisation.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
A real review surface is committed and the control structure around it is not published. The vendor states that every summary is reviewed by legal professionals before it reaches the customer and lists humans in the loop as one of three trust pillars, which is a documented gate on every output rather than an aspiration. Section 6 of the governing terms adds a second checkpoint by making the customer solely responsible for verifying accuracy and completeness before taking or omitting any action. What is missing is everything around those two points: no threshold is published at which the system defers, nothing describes what the reviewing professionals check or against what standard, no reviewer qualification is stated, and nothing addresses what happens when an error reaches a filed brief.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
Real deployment evidence with substance, short of measurement. Eleven customer firms appear as logos including Mueller Law, Mann and Potter, Explico, Windsor Troy Law, the Hernandez Legal Group, Rencher Law Group, Amini and Conant, Struble Cohen Trial Lawyers, Broussard Knoll and Conforto Law Group. Four testimonials carry names and are tied to firms rather than floating free: Wolfgang Mueller of Mueller Law, Delaney J. Miller of Windsor Troy Law, Matthew Struble of Struble Cohen who identifies himself as board certified in both civil trial and appeals, and a reviewer at Conant. The quotations contain operational detail a reader can weigh, including a stated processing time of about 30 minutes for a transcript and a note that only minimal adjustments were needed on review. What is absent is measurement: no figure for time or cost saved is attached to any named firm, nothing is dated, and a customers page exists that was not opened in this pass.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
Confidentiality is addressed in general commercial terms and nothing addresses client confidences as such. Section 9 of the governing agreement is a conventional mutual confidentiality provision: customer content is the customer's confidential information, the recipient must use reasonable care, and disclosure is limited to personnel and subcontractors under equivalent obligations. That is readable in advance, which keeps this off the floor. Everything the A band asks for beyond it is missing. Privilege and work product are never mentioned, on a product whose entire input is deposition testimony. Nothing addresses segregation between customers or matters. No retention period is stated and section 2.2 instead reserves to the vendor the right to set the maximum period it will retain customer content. Training is permitted rather than prohibited under section 4.2. One provision cuts against the product's own market: section 2.3 bars submitting protected health information or sensitive personal data unless expressly authorised in an Order, which is a live constraint for medical malpractice and personal injury depositions.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point.
Nothing published addresses the advice line, and this rests on a document read in full rather than on an untested gap. The governing terms of service were retrieved on 4 September 2026 and contain no statement that the output is not legal advice, no competence or supervision language, no reference to any rule of professional conduct, and no jurisdiction limit. Section 6 requires the customer to verify accuracy before acting, which is framed as an accuracy precaution rather than a professional responsibility one. The marketing describes the product as built by attorneys for attorneys and founded by a practising civil rights attorney, and the output is explicitly intended for copying into motions and briefs and for submission to the courts, which raises the professional responsibility question rather than answering it. The one adjacent provision is the site's own claim that every summary is reviewed by legal professionals, which is a quality control statement and not a position on advice.
AI Governance and Bias Disclosure
Published governance over model behaviour: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
No governance position is published for a system whose output is copied into court filings. There is no responsible AI statement, no governance framework, no named owner, no description of pre-release testing, no evaluation results and nothing whatever on bias. The nearest published material is section 4.2 of the agreement, which describes where the models sit and who may see them, stating that Rev's speech recognition and other AI models are proprietary, maintained locally and not shared with any third party; that is a data handling statement rather than a governance regime. Searched the home page, the pricing page, the terms of service and the site navigation on 4 September 2026.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
The governing agreement covers the product without addressing what happens to transcripts after processing. Retention is acknowledged and then left open: section 2.2 reserves to the vendor the right to establish general practices and limits including the maximum period it will retain customer content, and no period is published anywhere. No deletion commitment is stated. Security is addressed only at the level of section 9.3, which promises measures in accordance with industry standards and applicable law, with no controls named. No incident or breach notification practice was located. Subcontractors are acknowledged in section 1.3, including individual freelancers who perform human-based services, with none named. A data processing addendum and a business associate agreement are incorporated by reference at the parent's domain and were not opened. Section 2.3 places the backup obligation on the customer and excludes vendor liability for loss of data.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
A real published position, carried almost entirely by one clause most vendors omit. Section 10 commits the vendor to maintain insurance at stated minimum limits for the term, including commercial general liability at one million dollars per occurrence and, decisively for this axis, professional errors and omissions cover for the Service including network security and data protection liability at not less than two million dollars per claim and four million in the aggregate. That is the insurance limb this band's top rung asks for and no other record in this pull publishes it. Against that, the rest is thin. The indemnity at section 7 runs only from customer to vendor; there is no indemnity to the customer for anything, including third-party infringement. Section 8 caps liability at fees paid in the preceding twelve months for the services subject to the claim. Section 6 disclaims all warranties and accuracy expressly, and while it carves out a warranty said to be at section 6.2, no such subsection appears in the published document, so the one promised warranty cannot be read. No insurance certificate or carrier is named.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
No integration into practice systems was located. No document management system, case management platform, transcript repository, word processor or e-signature product is named anywhere on the surfaces read, and there is no integrations page or developer documentation in the navigation or footer. The only connectivity evidence is indirect, in section 1.2 of the agreement, which refers to APIs made available to facilitate use of the services without describing what they connect to. The workflow the product describes is manual: upload a transcript, receive documents, copy and paste citations into briefs. A Court Reporting Solutions page exists in the Products menu and was not opened in this pass; it is named here so the limit is visible rather than presented as settled. Checked 4 September 2026.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
Nothing is published about where the software runs or where transcripts sit. No region, country, hosting provider or data centre is named on any surface, and no tenancy model is described: nothing states whether the platform is single or multi-tenant or how one customer's transcripts are separated from another's. The agreement is silent on both, referring only to Rev's servers in the context of storage allocation limits at section 2.2. There are no deployment options, no private or dedicated tier, and no residency commitment of any kind. Searched the home page, the pricing page, the product overview, the terms of service and the footer on 4 September 2026.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
No independent security attestation was located. No SOC 2, ISO, HIPAA certification or any other framework is claimed anywhere on the site, there is no trust centre or security page, no auditor is named, and no report is offered at any access tier. The site carries no security badges at all, so there is nothing unsupported on display either, which is why this sits at the floor rather than in the badges-without-scope band. The agreement offers only section 9.3, promising security measures in accordance with industry standards and applicable law without naming a standard. A published customer testimonial refers to extensive discussion with the team on security before purchase, which indicates that security assurance here is a sales conversation rather than a published artifact. Checked home page, pricing page, terms of service and site footer on 4 September 2026.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
The supply chain is partly disclosed, in two fragments that sit on different surfaces. The home page carries an icon labelled OpenAI Zero Data Retention among its three trust pillars, which names a third-party model provider and states the retention posture agreed with it, though no supporting text explains scope or which features it covers. Section 4.2 of the agreement describes the other half of the architecture: the vendor's own speech recognition and generative models, stated to be proprietary, maintained locally and not shared with or disclosed to any third party. So a reader can establish that both a named external provider and proprietary internal models are in use. What is absent is the rest: no specific model or version is named, no location is given for any of it, no subprocessor or model provider list is published, and no commitment to notify customers when a model or provider changes was located.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
A buyer can price the product completely without speaking to anyone. Two purchase routes are published side by side. The platform subscription is 99 dollars per user per month and reduces summaries to a flat 25 dollars each at any volume, with the included capabilities itemised as AI deposition chat, automated memos, contradiction finder, and theme extraction and clips. Standalone summaries are published as a volume table with five bands: 85 dollars for one to nine, 75 for ten to nineteen, 65 for twenty to thirty, 55 for thirty-one to forty, and 50 for forty-one and above. An interactive estimator on the page computes a monthly total against the pay-per-summary alternative and shows the saving. A seven-day free trial is offered with no credit card required and cancellation at any time, and signup is self-serve. The unit is unambiguous throughout, being one summary per deposition transcript. What is not published is any figure for implementation, since none is required.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
Coverage is asserted at the level of a whole profession without practice detail or boundaries. The buyer is described as attorneys and litigators generally, with the marketing framed as built by attorneys for attorneys, and a second segment is addressed through a Court Reporting Solutions page for court reporting agencies, which was not opened. No practice area is named anywhere as supported: the named customer firms span trial and personal injury practices but the vendor itself claims no area of specialisation, no matter type and no transcript type. Firm size is not addressed, in-house and government use are not mentioned, and nothing states where the product stops. The product is narrow by nature, covering deposition transcripts only, but that scope is implied by the offering rather than stated as a limit.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The published agreement expressly reserves a right to train on customer content, with no opt out located. Any de identification, anonymisation or aggregation qualifier is recorded in the summary.
Section 4.2 of the governing agreement states that if the customer subscribes to any of Rev's speech-to-text services, customer content will be analysed by Rev's speech recognition models and other Rev artificial intelligence models and may be used for continuous training of those models. The clause names training expressly and operates on customer content, which is what this value turns on. Two qualifiers are recorded rather than treated as removing the permission: the same clause states that customer content will not be used for any generative AI model training, and that Rev's models are proprietary, maintained locally and not shared with any third party. The home page carries an icon reading OpenAI Zero Data Retention, which addresses the third-party generative provider and is consistent with the generative carve-out; neither statement displaces the permission to train Rev's own models. The condition attaching the clause to speech-to-text services is left as the agreement states it, since the agreement never describes deposition summarisation.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Retention is acknowledged in public materials with no stated period.
Retention is acknowledged in the agreement and no period is published. Section 2.2 provides that the vendor may establish general practices and limits concerning use of the platform and the services, including the maximum period it will retain customer content and the maximum storage space allotted, which reserves the question to the vendor rather than answering it. No default window, no deletion commitment and no certification of deletion appears anywhere. Section 2.3 places the backup obligation on the customer and excludes vendor liability for any loss of data. Searched the home page, the pricing page, the terms of service and the site footer on 4 September 2026.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
No located public material addresses walls or matter level segregation.
No located public material addresses walls or matter level segregation. Nothing states whether the platform is single or multi-tenant, how one customer's transcripts are separated from another's, or how access is controlled within a customer account, and no role or permission model is described. The agreement's only adjacent provisions are access restrictions on the customer's own authorised users at section 2.2 and a confidentiality obligation on personnel and subcontractors at section 9.2, neither of which describes segregation of stored content. There is nothing to quote because the position is absent rather than asserted. Searched the home page, the pricing page, the product overview and the terms of service on 4 September 2026.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
Section 9.2 of the agreement provides that if either party receives a court subpoena, a request for production of documents, a court order or a requirement of a government agency to disclose confidential information, the recipient will give prompt written notice to the other party so that the request can be challenged or limited in scope. Customer content is expressly defined as the customer's confidential information at section 9.1, so the commitment reaches deposition transcripts. The clause is mutual and is not qualified by a where legally permitted carve-out, which is unusual. No transparency report of such requests was located on any surface.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
No located public material identifies the corpus behind the product’s answers.
No located public material identifies a corpus behind the product's output, and the product's design makes the question narrow: SmartDepo summarises the deposition transcript the customer uploads rather than retrieving external legal content. No database, publisher, jurisdiction or licence basis is named anywhere, and the citations the product generates point to page and line numbers within the customer's own transcript. Searched the home page, the pricing page, the product overview and the terms of service on 4 September 2026.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
Nothing on any located surface addresses whether authority is checked for subsequent history. The product does not retrieve primary law at all: it reads a deposition transcript and produces summaries, admissions analyses, abstracts and memos citing back to page and line within that transcript. The question therefore does not bite on this product class and the honest value is the absence rather than a penalty. Searched the home page, the pricing page, the product overview and the terms of service on 4 September 2026.
Refusal and Uncertainty Behaviour
What does the product do when the answer is not in the corpus?
No located public material addresses what the product does when it cannot ground an answer.
No located public material describes what the product does when it cannot ground an output. The marketing runs the other way, promising complete coverage with nothing left out and no errors of omission, and the agreement acknowledges at section 6 that the services may contain errors while placing verification on the customer, but neither describes an abstention path, a no-answer state, or any confidence or grounding signal surfaced to the reviewer. The published human review step is a quality check on delivered output rather than a described behaviour of the system. Searched the home page, the pricing page, the product overview and the terms of service on 4 September 2026.
Fabricated Citation Record
Does a public court record exist involving output from this product?
No court order, opinion or disciplinary record naming this product has been located as of the date shown. This is a statement about the public record, not a finding about the product.
The AI Hallucination Cases database maintained by Damien Charlotin was searched on 4 September 2026 on both the product name SmartDepo and the parent company name Rev.com. No court order, opinion or disciplinary record naming the product was located. This records the state of the public record on that date and is not a finding about the product.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No located public material engages with bar or ethics guidance.
No located public material engages with bar or ethics guidance. No bar association, regulator, rule of professional conduct or ethics opinion is named anywhere in the agreement, the home page or the pricing page. The nearest references are marketing identity claims, that the product is built by attorneys for attorneys and founded by a practising civil rights attorney, which say who built it rather than which professional standards its use engages. This matters more than usual because the output is expressly intended for copying into motions and briefs and for submission to the courts. Searched on 4 September 2026.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Public materials claim time savings without addressing billing or disclosure.
Public materials claim time savings without addressing billing or disclosure. The home page states that the product saves weeks of manual work, and published testimonials describe processing in about 30 minutes, saving hours of time, and producing in minutes what previously took many hours. One testimonial refers to a reasonable pricing structure and another to obtaining page-line summaries being the most time-consuming aspect of trial preparation. Nothing addresses what happens to a client bill when that time disappears, and no per-matter record of AI-assisted work is described. The vendor's own charging model is per summary or per user rather than per hour, which is a cost fact rather than an answer to this signal.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
The material exists behind a sales conversation or an executed agreement.
The material a firm would need is referenced rather than published. A data processing addendum and a business associate agreement are incorporated into the agreement by reference at section 9.3 and located at the parent's domain rather than on any SmartDepo surface; neither was opened in this pass. No subprocessor or model provider list is published anywhere. Section 1.3 acknowledges that subcontractors including individual freelancers perform parts of the services without naming any of them, which is material a client would want identified given those individuals handle deposition testimony. The only provider naming located is an icon on the home page reading OpenAI Zero Data Retention, with no supporting text stating scope or which features it covers.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Some elements of the record are available, short of a document level export.
Substantial elements of a record are produced, short of anything covering the AI itself. Summaries carry page-line citations hyperlinked precisely to the source dialogue, marketed as copy-and-paste ready for motions, briefs and submission to the courts, which covers the sources-retrieved element better than most records on this signal. Every summary is stated to be reviewed by legal professionals, which is a human verification step, though nothing describes it as recorded or exportable. What is absent is the model dimension and the export: nothing states that the model behind a given summary is recorded or disclosed, no per-document verification record is described, and no disclosure guidance or template for a court was located.