OurFirm.ai: Revenue, team and how it was built
OurFirm.ai (AI litigation platform for trial lawyers: drafting, citation checks, discovery review and judge profiles, United States): US$21K MRR reported to Antler, April 2026. Evidence grade D: reported, not verified. Team: 1 founder, trial attorney Andrew Mancilla, who built it in 2025; a ~10-person team by November 2026 (careers page). Launched October 2025; still running.Added 5 November 2026 · updated 6 November 2026
- Outcome
- US$21K MRR reported to Antler, April 2026
- Evidence
- D reported, not verified
- Launch
- October 2025
- Status
- Still running
- Team
- 1 founder, trial attorney Andrew Mancilla, who built it in 2025; a ~10-person team by November 2026 (careers page)
- Outside funding
- Antler portfolio company; US$4M total funding listed by Stanford CodeX TechIndex
- Group
- Legal and law firms
- Main channel
- Cold outreach
What is not verified
The headline figure of US$21,000 in MRR rests on one statement: Antler's Spring 2026 U.S. showcase post of 8 April 2026, written from the company's own numbers. Antler is an investor in the company. The Agent Jesse list of August 2026 names the company in the same Antler cohort and is not a second source. The US$4 million funding total is a directory value from Stanford CodeX TechIndex, without a press release. No payment-processor profile, filing or later dated revenue figure was found. A reader could look for a funding announcement or a later interview with the founder. The headline number is grade D: one claim or one estimate, with no second source. Aiolum shows it because the story is useful, and marks it so that it is not mistaken for a verified figure.
OurFirm.ai is a litigation platform: software for lawyers who argue cases in court, as opposed to lawyers who write contracts. It uses large language models (AI models that read and write text) from outside providers, arranged as an agent swarm: many small AI programs that each do one part of a task, such as research, drafting or citation checking, at the same time. A feature called Check My Cites confirms that each cited case exists, that the quotation is accurate and that the case is still good law. Judge AI builds a profile of a judge from opinions, transcripts and docket entries.
Andrew Mancilla, a trial attorney admitted in New York, New Jersey and California, founded it in 2025, and the founder page says he built it inside his active trial practice. On 8 April 2026 the investor Antler wrote in its Spring 2026 U.S. showcase that the company "Grew from $0 to $21K MRR entirely through word-of-mouth". MRR means monthly recurring revenue, the monthly value of current subscriptions. By November 2026 the careers page described a "~10-person team" and Stanford's CodeX TechIndex listed US$4 million in total funding.
The revenue figure is one statement, published by an investor in the company, so this page grades it D: reported, not verified. The page also shows earlier legal-technology companies that failed with a similar promise, including an AI legal research company.

How much revenue does OurFirm.ai make?
Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.
| Fact | Value | Grade | How we know | Sources |
|---|---|---|---|---|
| Monthly recurring revenue, April 2026 | US$21,000 | D | Antler's Spring 2026 U.S. Portfolio Showcase post, 8 April 2026: 'Grew from $0 to $21K MRR entirely through word-of-mouth'. A founder figure published by an investor. The Agent Jesse list of August 2026 names the company in the same cohort and is not a second source. | [1, 2] |
| Yearly run rate of that MRR | US$252,000 | calc | Calculated: US$21,000 times 12. | [1] |
| Daily active users, April 2026 | 35%+ daily active user rate | D | Same Antler sentence: 'attorneys opening OurFirm before Westlaw'. It does not say 35% of which group. | [1] |
| Founder | Andrew Mancilla, founder and CEO | C | Home page: 'Andrew founded OurFirm.ai in 2025'. Antler names 'Andrew (CEO)' and no co-founder. | [3, 1] |
| Founder background | trial attorney, J.D. Brooklyn Law School 2010 | C | The company's pages say 15 years of trial practice; Antler says '14-year trial attorney' who ran his own boutique litigation firm. | [7, 1] |
| Team, November 2026 | about 10 people | C | Careers page: 'Stage signals: ~10-person team'. The about page names three people: the founder, a head of sales and a CTO. Five open positions were listed. | [6, 5] |
| Funding | US$4M total, per directory | D | Stanford CodeX TechIndex lists 'Total funding $4M'. Antler's post presents it as a portfolio company. No press release with the round was found. | [11, 1] |
| Company | Ourfirm.ai, Inc., Delaware law, headquarters New York | C | Terms of service name the company and Delaware law; the about page says 'Founded 2025 · New York'. | [10, 5] |
| First dated public document | 23 October 2025 | C | Subprocessor list, 'Effective Date October 23, 2025'. The launch month is not published; this page uses October 2025 as the launch date. | [8] |
| Time from first dated document to the figure | 6 months | calc | Calculated: October 2025 to April 2026. | [8, 1] |
| Prices, November 2026 | US$150 and US$600 per user a month, plus usage | C | Pricing page, read 5 November 2026. | [4] |
| AI model providers | OpenAI, Anthropic, Google Gemini | C | Subprocessor list, last updated 24 July 2026. The about page names 'Anthropic Opus and the latest frontier models'. | [8, 5] |
| Conferences attended, 2026 | ILTACON, Legal Innovators, Legal Tech Talk London, CLOC and others | C | Events page lists past and upcoming events from April 2026 to 2027. | [9] |
How to read this table: grade D is a single statement that no one outside checked, or a directory value. Grade C is the company's own website. Grade X was computed on this page. No value is grade A or B.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Professional | 150 | per user, plus usage; 20M token context window, drafting, citation checks, Judge AI | |
| Enterprise | 600 | per user, plus US$600 one-time prepaid usage; 30M context, client portal, custom builds |
Both plans add usage charges on top of the seat price; the page says 'a quiet month costs little and a heavy litigation month scales with the work'. Fees are paid in advance, monthly or annual; the page gives no separate yearly price. Volume licensing and custom contracts are by request.
When did OurFirm.ai launch, and how did it grow?
From a trial attorney's own practice in 2025 to a reported US$21,000 a month in April 2026 and a team of about ten. Each entry carries its grade.
- 2010CAndrew Mancilla receives his J.D. from Brooklyn Law School and starts a trial practice in criminal, commercial and civil litigation. [7]
- 2025CHe founds OurFirm.ai and builds it inside his active trial practice. [7, 3]
- 23 October 2025CThe subprocessor list is first published, naming AWS and the AI model providers. It is the earliest dated document on the site. [8]
- 8 April 2026DAntler's Spring 2026 U.S. showcase reports US$0 to US$21,000 in MRR through word of mouth and a 35%+ daily active user rate. [1]
- April to October 2026CThe company shows the product at legal-technology conferences and CLE panels, including ILTACON in August 2026. [9]
- July to September 2026CThe blog publishes posts on Judge AI, AI sanctions and the Heppner ruling, and announces a SOC 2 Type 2 attestation on 16 September 2026. [7]
- November 2026CThe careers page describes a ~10-person team and lists five open sales and marketing roles in New York. [6]
Why did OurFirm.ai work?
Four causes, in the order they mattered. Each one is a reusable lesson.
A practicing lawyer as the product designer
The company says every prompt, workflow and template was written by the founder, a trial attorney. That knowledge of court documents, deadlines and judges is the part a general AI company does not have.
One narrow kind of legal work
The product serves only contested litigation: motions, discovery, trial preparation and judges. Its comparison page sets it against general legal AI tools such as Harvey and Legora, and against general chat assistants.
Checks against invented citations
Courts have sanctioned lawyers who filed AI text with invented cases. The product runs every citation through a multi-step check and the blog writes about these sanctions, so verification is part of the sales message.
Usage billed to the lawyer's client
The pricing page says the usage charge is built to pass through to clients as a disbursement, with a client-savings report for each matter. The firm does not carry the AI cost as overhead.
How it was built
These details come from the company's own pages. They are grade C. [8, 3, 5, 4]
- Language models from OpenAI, Anthropic and Google Gemini, named on the subprocessor list. Mistral AI was listed in October 2025, and Cerebras was removed in July 2026 as never engaged.
- Reducto.ai for OCR, the conversion of scanned pages into text.
- Amazon Web Services for compute and storage, described in the FAQ as AWS EKS (Kubernetes) in several availability zones.
- An agent swarm: a specialized sub-agent for each legal issue, running in parallel, and a context window of 20 million tokens on the Professional plan.
- Its own data: judicial opinions, transcripts and docket entries, with semantic search over PACER (federal court records) and NYSCEF (New York State court filings).
- Integrations with SharePoint, Dropbox, Box, ShareFile and Clio, and security audits (SOC 2) that law firms ask for.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before OurFirm.ai, and why did they fail?
Four companies in our database of failed startups tried to sell technology to lawyers and failed. The values below are from that database and are not independently verified.
ROSS Intelligence
The promiseAI legal research that answers questions in plain English.
Why it diedIt could not win customers from Westlaw and LexisNexis, which held decades of data, and law firms were slow to change.
Atrium
The promiseA technology-driven law firm for startups, with automated legal work.
Why it diedThe automation could not handle real legal complexity, so it carried the costs of a law firm without its quality.
jEugene
The promiseAutomated checking of contracts for common errors.
Why it diedBetter-funded competitors with stronger AI arrived, and law firms adopted slowly.
Glider
The promiseOne place to manage and negotiate contracts.
Why it diedIt could not set itself apart from DocuSign and Adobe Sign.
The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.
What changed
ROSS Intelligence and jEugene had to build their own language technology for legal text, with years of work and little money, and lost to companies that held more data. OurFirm.ai rents its language models from OpenAI, Anthropic and Google, so the hard part that remains is legal knowledge, court data and checking, which a practicing lawyer can supply.
Atrium failed because software could not do the legal work by itself. OurFirm.ai sells to lawyers, not to their clients: the lawyer reviews and signs every filing, and the product checks citations so that the lawyer can verify what it drafted. The founder's own practice was the first test case, and Antler's post reported that early growth came from word of mouth among attorneys.
Can one person copy OurFirm.ai now?
Only with the founder's profession.
The model access is the same for everyone. The part that is hard to copy is fifteen years of trial work, the founder's network of New York litigators, and court data collected over time. A professional in another field, such as tax, insurance claims or immigration, can copy the structure: build the tool for one's own practice first, then sell it to colleagues.
Where it is still open: State courts beyond New York, which the FAQ says are still in the pipeline, and other professions that file documents with strict rules.
- 1Pick one kind of professional work that you do yourself and that has strict rules, such as court filings.
- 2Build the tool on rented models, with your own templates and prompts for that work.
- 3Add a check that verifies every factual claim against the source, because errors are costly in that field.
- 4Sell to colleagues one by one, with a price that the professional can pass on to the client.
What could go wrong for OurFirm.ai?
The risks are the mirror image of the causes above.
What is not verified
The headline figure of US$21,000 in MRR rests on one statement: Antler's Spring 2026 U.S. showcase post of 8 April 2026, written from the company's own numbers. Antler is an investor in the company. The Agent Jesse list of August 2026 names the company in the same Antler cohort and is not a second source. The US$4 million funding total is a directory value from Stanford CodeX TechIndex, without a press release. No payment-processor profile, filing or later dated revenue figure was found. A reader could look for a funding announcement or a later interview with the founder.
Team size
The founder started alone in 2025. The careers page in November 2026 says '~10-person team' and lists five open roles; the about page names three people. No source gives the headcount in April 2026, when the revenue figure was reported.
Judge AI
The home page calls Judge AI a 'Frontier proprietary emulation of federal judges - predict how your judge will rule before you file a single page', and the FAQ says it creates 'interactive behavioral AI profiles of judges trained on their actual transcript and opinion data'. The site also publishes profile pages for named New York judges. The sources do not discuss how judges or courts view this use of their records.
Testimonials and claims on the home page
The home page shows quotations from named attorneys and firm names, and says 'Trusted by hundreds of elite litigators' and 'Zero hallucinated citations have been filed by OurFirm.ai users'. These are the company's own statements; no outside source confirms the count of users or the citation record.
Competition
OurFirm.ai competes with other legal AI stories on this site, such as Legion AI Associate for litigation drafting and Descrybe for legal research, and with large funded companies that its own comparison table names, Harvey and Legora.
Dependence on model providers
The subprocessor list names OpenAI, Anthropic and Google Gemini for model inference, and the about page says it uses 'Anthropic Opus and the latest frontier models'. A price change or a change of terms by these providers would change the cost of every plan, although the usage charge passes part of that cost to customers.
Sources
The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.
- D Antler, 8 April 2026: Announcing the Spring 2026 U.S. Portfolio ShowcaseRead the section on OurFirm.ai, downloaded on 5 November 2026 after reading robots.txt (it allows /blog/ and asks for a 10-second crawl delay, which was respected; the blocked /portfolio/ pages were not used). Source for the MRR, the daily active user rate and the founder's background.
- D Agent Jesse, August 2026: Antler 2026 cohort companiesRead the list in the harvester's copy of 6 November 2026. Names OurFirm.ai in the Antler 2026 group; no new figure.
- C OurFirm.ai home page and FAQ, read 5 November 2026Read in full. Source for the features, the FAQ on security, data and prices, Judge AI and the testimonials.
- C OurFirm.ai pricing page, read 5 November 2026Read in full. Source for the plans and the usage billing.
- C OurFirm.ai about page, read 5 November 2026Read in full. Source for the founding year, the named team and the models used.
- C OurFirm.ai careers page, read 5 November 2026Read the list of positions and the Head of Marketing posting. Source for the ~10-person team.
- C OurFirm.ai founder page: Andrew Mancilla, read 5 November 2026Read in full. Source for the founder's background, the 2025 launch and the dated blog posts.
- C OurFirm.ai subprocessor list, effective 23 October 2025, last updated 24 July 2026Read in full. Source for the AI model providers, AWS, Reducto and the change log.
- C OurFirm.ai events page, read 5 November 2026Read in full. Source for the conferences attended.
- C OurFirm.ai terms of service, effective 1 March 2026Read the sections naming the company and the governing law.
- D Stanford CodeX TechIndex: OurFirm.ai, read 5 November 2026Read in full. A directory entry; source for the founding year and the US$4M total funding.
- C Database of failed startups, local exportLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.
Run this company? Send corrections or a verified figure to aiolum@livyden.com