AfterQuery Data Partnerships: What It Buys and Pays

AfterQuery's data partnerships license an established company's records of how its teams work, from email and chat to CRM and finance, to train and evaluate frontier AI, and take operating, winding-down and closed companies alike. It publishes no price. It prices by the depth, breadth and uniqueness of the records and proposes terms after reviewing a sample, and its application form asks you for the price you expect, in bands from under $50k to more than $500k. It removes personal details itself: names and identifiers are removed or replaced with consistent pseudonyms, and every dataset is checked with a re-identification test before delivery. AfterQuery is a well-funded buyer, reported in September 2026 at a $3.2 billion valuation. What its pages do not say is what you would be agreeing to: no licence length, exclusivity or right to pass the data on is published, and payment is the last of its six steps, with no schedule or number of days.

This guide is for the founder, CEO, CFO or counsel of a company that has searched for AfterQuery by name, heard it pays for company data, or been told someone wants to introduce it, and wants to know what it buys, what it might pay a company like theirs, whether they qualify, what they would be agreeing to and how to apply or be introduced. It reads AfterQuery's own pages, its terms, its job posts and the press about it, in the order a company meets them: what AfterQuery is and why it buys, what it licenses in each industry, who it takes, how it prices, its six steps, what happens to personal data, how it pays, the licence, and how to apply or be introduced. It ends with a comparison against Mercor, micro1, Handshake AI, Turing and Scale AI, and a worked example.

AIData.Broker is in AfterQuery's referral programme, as it is in Mercor's, Handshake AI's, Turing's, micro1's and Scale AI's. We introduce a holder to every buyer whose published rules take its data, free for the holder, because the buyer pays us its published referral payout out of its own money: for AfterQuery, up to $85,000 for a company it goes on to license data from, the most its data partnerships page names, and the holder sees the most any buyer may pay us before we introduce it. Where a sale run by us earns more, and only for holders based in the United States, the United Kingdom and Canada, we act as the holder's agent under a mandate and the holder pays our fee only if the sale closes; every fitting partner is introduced then as well. Either way the holder itself licenses its data to the buyer, and we do not buy it, take ownership of it, or license or sell it in our own name. Every term below is AfterQuery's own published wording or named press, cited to its record; where a line is our reading, it says so.

This guide is research from public sources, not legal, tax or investment advice.

Contents

1. What AfterQuery is, and why it buys company data
2. What AfterQuery licenses, industry by industry
3. Who it takes
4. How it prices, and what it might pay
5. The six steps, from form to payment
6. What happens to personal data
7. How and when it pays
8. The licence: what is published and what is not
9. Applying directly or being introduced
10. AfterQuery beside Mercor, micro1, Handshake AI, Turing and Scale AI
11. Worked example: a freight brokerage past its peak
12. Questions people ask
13. What this means for you

What AfterQuery is, and why it buys company data

AfterQuery is a San Francisco company, founded in 2025 by Spencer Mateega and Carlos Georgescu, that sells expert-built training data, RL environments and custom evaluations to frontier AI labs and enterprises. It calls itself "an applied research lab curating data solutions for frontier foundation model development" ( its Y Combinator page). Much of what it sells is made with people it pays: its expert data is, in its words, "captured from domain experts step by step, decision by decision, and verified before delivery", and it also generates synthetic data and helps "turn messy enterprise data into a form models can be trained on" ( its solutions page). Its data partnerships are a newer line, which licenses records a company already holds.

It has grown fast. It raised a $30 million Series A at a $300 million valuation, led by Altos Ventures with The Raine Group, in April 2026, and said the same month that it had surpassed $100 million in revenue run rate, a figure that is its own. In September 2026 it reportedly raised a round that valued it at $3.2 billion, which TechCrunch reported as making it Y Combinator's fastest-ever unicorn. A run rate and a valuation say how large and well funded the business is. They do not say how much it spends on company records, which it does not publish.

Why it wants your records is on its data partnerships page: "AI models are learning to do real jobs. To get better, they need a record of how real teams work. Who asked for what, what was decided and how it turned out. Your company has been building that record for years, and it can't be recreated" ( data partnerships page). Labs train agents to work inside companies by letting them practise in RL environments, simulated workplaces where a model attempts a task and is scored on the result, and AfterQuery is hiring software engineers to build those environments. A company's real email threads, tickets, CRM history and finance workflows are what makes such a workplace realistic. Buyers say public data and synthetic datasets cannot replicate that, which is why they pay companies for the real thing.

It also has a person whose job is to buy. Its Data Acquisition Lead, by its own job post, owns "how we find and reach the companies that own valuable real-world data, how we price it, and how we negotiate and close those purchases". A company that talks to AfterQuery is talking to a buyer that has staffed the pricing and negotiation of exactly this kind of deal, and should prepare accordingly (see How it prices).

AfterQuery is a principal: it licenses your records directly from you, for its own account, and it says it uses them to train and evaluate frontier AI. Our reading: since AfterQuery sells datasets, environments and evaluations to labs rather than training models of its own, your records become the material for what it sells to them, and its page does not say whether labs receive the records themselves or only what AfterQuery builds from them. That is the first question for the licence (see The licence). The programme is new, and no company has yet confirmed in public a licence of its records to AfterQuery.

What AfterQuery licenses, industry by industry

AfterQuery's answer to "What kind of company data do you license?" is short: "Records of how teams actually work across their tools: email, chat, tickets, CRM, support, finance and more, usually over several years. Industry systems and legacy tools count too" ( data partnerships page). The tools its page shows are the everyday ones: Slack, Microsoft Teams, Gmail, Google Drive, Jira, HubSpot and Salesforce. It wants workflow records, the messages and documents a business produces by working, not a curated dataset, and its words "legacy tools count too" mean an old ticketing system or a retired CRM export is in scope.

The nine industries on its page

Its page lists nine industries, each with the work it wants from them. In its own words ( data partnerships page):

Read the list as a description of demand, not a rule. It is broad (it covers most companies that run on software), and nothing on the page turns a company away for being in an industry it does not name. What the nine have in common is the kind of record AfterQuery is after: a piece of skilled work with its steps and its outcome, such as a claim and how it was handled, an RFI and the change order it produced, a month-end close from first reconciliation to sign-off, or an incident from the alert to the fix. That is what an AI model can learn a job from.

Two of the nine carry legal duties before they carry a price. Healthcare names de-identified clinical records, and in the US a record is outside the HIPAA Privacy Rule only once it meets its de-identification standard; a practice should read HIPAA de-identification for AI training before any clinical note leaves it. Legal says privileged material is never used, which is a promise about AfterQuery's use, not a decision the firm can hand over: the rule that rescues an inadvertent disclosure of privileged material ( Fed. R. Evid. 502(b)) does not cover a deliberate commercial transfer, so the firm decides what is privileged and keeps it out before export, as the Spirit Airlines records sale did when it excluded privileged material in the agreement itself.

What makes records worth more to it

AfterQuery's price factors (below) favour records that link across systems: a ticket, the chat thread where the fix was argued and the email to the customer who reported it. Other buyers say the same. The head of Sunset, which buys closing companies' records, told Forbes that a Jira ticket tied to a specific code commit is worth more than a standalone document, and micro1's chief revenue officer said that without Slack, CRM or email records the rest of a corpus is worth less. So the systems a company is most nervous about sharing, its chat and its email, are usually the ones that carry the value. That is a reason to decide early, and carefully, what inside them stays out, not a reason to share them whole. Our pages on Slack, Microsoft 365, Google Workspace, HubSpot, Salesforce and Atlassian set out what each system's own terms allow.

Who it takes

AfterQuery states its rule in one answer. "Does my company qualify?" "We're most interested in established companies with several teams working across several systems, and a few years of history. Operating, winding-down and closed companies are all welcome." It states no number, no country and no language.

Size and years: our reading of "several teams" and "a few years"

We read those words as a rule, and we read them this way: at least 10 people, counted at the company's largest, and at least 3 years of history ( AfterQuery's criteria as we hold them). The numbers are ours, not AfterQuery's: "several teams" is more than one team of a few people, and "a few years" is more than two. We count the people at the company's largest because AfterQuery's words describe the teams that made the records, not the staff on the payroll today, so a company that had several teams across several systems for years and has since shrunk still holds the record those teams made. AfterQuery's own form does not exclude small companies either: its smallest size band is one to ten employees ( its application form). So a company below our reading is not ruled out by anything AfterQuery wrote; its own team decides at review, and such a company should expect to make its case on what its records contain.

The reading puts AfterQuery among the most open of the large buyers on size. Handshake AI looks for 20+ full time employees and 3+ years in operation, micro1 for 30+ full-time employees, Scale AI says 40+ employees is its preferred starting point, and Mercor sets a size for each kind of company on its listings, the smallest from five people. Only Turing names no size at all. And AfterQuery's "several systems" matters as much as its "several teams": a company whose records sit in one tool has less of what it prices (see How it prices).

Operating, winding down or closed

All three are welcome, in AfterQuery's words, as they are at Turing, which takes active, winding-down and wound-down companies through the same process. For a company still trading, the practical questions are scope and staff: which teams and systems to include, and what employees are told. For one that is closing, time comes first: records lose value fast once subscriptions lapse and admin access goes (Turing's intake puts it at 30 days for Slack and 60 to 90 days for a CRM after operations cease), so export first and decide later.

For one that has closed, the question is who may sign. A dissolved company's records still belong to someone: in Delaware, a dissolved corporation continues for three years to wind up its affairs and dispose of its property, after which a court-appointed trustee or receiver is needed, and a licence may also need the board and stockholder approvals the charter requires. A former CEO who still has the Slack export does not automatically have the right to license it. Read who owns a dissolved company's data, how long to keep business records after closing and our page for closing companies and estates before applying.

Where the company is based

AfterQuery's pages name no country, language or tax rule for a company that licenses to it. It is one of 14 buyers of company records whose own pages say nothing about where a holder can be based, and its website terms choose Delaware law. Silence is not a promise either way. A company outside the US should ask at its first contact whether AfterQuery takes companies from its country now, and plan for the cross-border questions: a US buyer may withhold tax on a licence fee paid abroad ( US withholding tax on data licence payments), and the records may be governed by privacy law the buyer must also meet ( selling data from outside the US).

The company's records, not its staff's

AfterQuery pays individuals as well as companies, and the two can meet in one person. Its expert programmes pay contributors per approved task, and one of its postings paid $300 per codebase a contributor connects and uses as a basis for tasks, for a private codebase the contributor owns. An engineer's work for a company is not theirs to connect: code an employee writes in the course of the job is a work made for hire, and the employer is its author and owns the copyright unless a signed writing says otherwise. Buyers have been offered records by people with no right to them ( one was offered a large company's codebase for millions, an illicit copy shopped to every data vendor), and they check title for that reason. A company that wants its records licensed should license them itself, as the company, and remind its staff that work material is not theirs to share.

How it prices, and what it might pay

AfterQuery's answer to "How is the price determined?" names four factors: "By depth, breadth and uniqueness: years of history, number of connected systems, the kind of work captured, and how hard it would be to recreate." It then proposes terms after reviewing a sample. There is no calculator, no listing and no figure on its pages. Each factor tells you what to prepare.

• Years of history. How far back each system goes, and whether the history is continuous. Its form asks how far back the records go, and its page says the records it licenses run "usually over several years".
• Number of connected systems. How many tools hold records of the same work, and whether they link (ticket numbers in commit messages, deal IDs in email subjects). Other buyers' estimators show how much this moves a price: Ooak's values a company with six connected tools at several times what it shows with three.
• The kind of work captured. Skilled work with its steps and outcome, in the workflows its industries list names. A support desk that records how each case was resolved is worth more than one that records only that it was closed.
• How hard it would be to recreate. A closed company's records cannot be recreated at all, and an operating company's years of decisions can only be recreated by living them again. This is the factor that makes old, specialist and legacy records worth something.

The price bands on its form

The one place AfterQuery mentions money to a company is its application form, which asks "What do you have in mind?" and offers bands from under $50k to more than $500k, beside "Not sure yet", with the note "A rough range is fine. It only helps us prepare" ( its application form). Our reading: whatever it is for, the band you pick is the first number in the negotiation, and it is given before AfterQuery has said anything about price. A company that picks a low band to seem reasonable may have set its own ceiling. "Not sure yet" is an honest answer until you have an independent reading of value; after one, answer from it.

For a sense of scale, other large buyers publish ranges for whole companies: micro1's tiers start at $100K-500K for a qualified partnership, Handshake AI's page says $100K to $4M+, Scale AI gives an illustrative $10K to $1M+, and Mercor's listings run from $10,000 to $1,500,000 a company, by kind. These are ranges across many companies, not what yours is worth, and micro1's and Nyne's public estimators run the same formula, so two of the numbers a company sees are one anchor rather than two. A free valuation of what you actually hold, what company data is worth and how AI data buyers calculate offers give you a number of your own before AfterQuery gives you its.

Scope is a price lever

AfterQuery says you decide the scope: which teams, which systems and which time window, and many partnerships start with one department. That is a real lever, and other buyers offer it too ( license one department first and keep the rest for later). It cuts both ways. A first deal for one department lets you see how AfterQuery works before you license more; it also lets AfterQuery see your records before it prices the rest. So settle in the first agreement what a later scope would be paid, and whether a renewal or a new delivery is priced on the same basis. AfterQuery's last step is "Payment and ongoing partnership", and buyers of company records talk of refreshes and ongoing partnerships but none publishes a refresh price.

The six steps, from form to payment

AfterQuery's page sets out six steps, which it calls "Short, safe and predictable." Its FAQ adds what happens after the first form and how long the whole takes.

From first form to signed agreement it is usually a few weeks, depending on how quickly a sample can be shared and reviewed. The pace is mostly yours: the questionnaire takes an afternoon, the sample takes as long as it takes you to export it and decide what to leave out.

Three things in these steps deserve attention. First, the NDA comes before the questionnaire, which is the right order: at least one buyer's terms say that until an NDA is signed, what you submit through its site carries no confidentiality obligation beyond its privacy policy. Read the NDA before you answer, because what you describe in the questionnaire is what the sample and the price will rest on. Second, the sample comes before the terms. AfterQuery's form says nothing is shared until there's an NDA and a signed agreement, while its steps put a sample before the terms are agreed; ask whether the sample counts as shared, what it may contain, and that it is deleted if you do not proceed. Third, transfer and de-identification are one step, and they come after the terms: the scrub is done by AfterQuery, on its side, after the records leave you (see the next section).

What happens to personal data

A company's working records are full of people: staff, customers, suppliers, the person who sent a sick note in a direct message. AfterQuery's answer to "How do you protect personal and customer information?" is the most specific of its published terms: names, contact details and other identifiers are removed or replaced with consistent pseudonyms, customer identities are masked or excluded, and sensitive categories such as payroll, privileged legal material and credentials are never used. Every dataset is checked before delivery, including a re-identification test, "because no process removes all risk on its own" ( data partnerships page). And on who sees the data: only the AfterQuery team working on your partnership, under NDA, and personal details are removed before anything is used in a product.

That is a serious set of statements, and its own last sentence is the honest one. Three questions follow from it.

Who holds the identifiable copy, and for how long. The removal happens after the transfer, so AfterQuery's team receives the records with the names still in them. That is the common model (Mercor runs its own automated pipeline, and Turing removes personal data on its side, air-gapped), but it is not the only one: Scale AI says it anonymises inside your own systems before anything is extracted and releases data package by package, each with your sign-off. Ask AfterQuery where the identifiable copy is held, who on its team can open it, and when and how it is deleted, and put the answers in the agreement rather than relying on the page.

What consistent pseudonyms leave behind. A consistent pseudonym replaces "Maria Lopez" with the same code in every email, ticket and chat, which is what keeps the record useful: a model can follow one person's work across systems. It is also what keeps that person linkable, and enough context (a job title, a city, a project) can point back to them. For staff and customers in the EU and the UK, the European Data Protection Board's view is that pseudonymised data stays personal data, so the company still needs a lawful basis to license it ( licensing data for AI training under the GDPR). In California, data counts as de-identified only with all three of reasonable technical measures, a public commitment not to re-identify, and contracts binding every recipient. Ask to see the re-identification test's result for each dataset, and what AfterQuery does when it fails.

Whose decision it is. AfterQuery's scrub protects AfterQuery's use of the records. It does not decide whether you may hand them over. Customer records raise what your privacy policy promised ( is it legal to sell customer data); staff messages raise your employees' expectations and, in some places, their unions ( employee emails and chats in AI training); suppliers' and partners' documents raise your contracts with them ( shared channels and third parties' attachments are reviewed or left out). Who controls de-identification is also a priced term in this market: in the Spirit Airlines auction, one bidder offered more for the right to de-identify with its own tools than with an independent party, and the estate chose the independent scrub. A company that wants an independent scrub before transfer, or wants to remove the most sensitive material itself first, should ask for it before the terms are agreed; our de-identification standard and de-identification tool show what a holder can do before any buyer sees a file.

How and when it pays

AfterQuery publishes the order of events and nothing else about payment: payment is the last of its six steps, after scope and terms are agreed and the data is transferred and de-identified. No schedule, no trigger, no number of days, and no word on whether any part is paid at signature. Its website terms say fees for its services are specified in separate agreements, which is about what customers pay AfterQuery, not what it pays you. Everything about your payment is in the agreement it proposes after the sample.

Against the other large programmes, that is the least that is published. Mercor pays by wire on delivery, typically 2 to 4 weeks after signing; micro1 pays in full within 30 days of acceptance; Handshake AI ties payment to delivery milestones it sets out in writing before any data is transferred; Turing pays from each sale of your record to a lab. Ask AfterQuery the questions those terms answer: when each payment is due, what event triggers it (signature, transfer, or AfterQuery's own acceptance of the de-identified dataset), whether it can reject a delivery and in how many days, and whether anything is paid before the data moves. Payment after transfer means the records are out of your hands before you are paid, so the date and the trigger belong in the agreement. For every buyer's payment trigger side by side, read when AI data buyers actually pay.

The licence: what is published and what is not

AfterQuery names the use and nothing else: it licenses your records to train and evaluate frontier AI. Nothing published says how long the licence runs, whether it is exclusive, whether AfterQuery may sublicense (pass the data or the rights on), which labs may receive the records or what AfterQuery builds from them, what you promise about earlier licences of the same data, or what happens to the records and to anything built from them when the partnership ends. All of it sits in the terms it proposes after reviewing a sample.

Its general website terms are the only published terms that touch your data, and they are written for the website and its services, not for a data licence. Two clauses in them matter. On ownership: "Any data or materials you provide to us remain your property, subject to the licenses granted to us necessary to provide our services". That is the right starting point, and the data agreement should say the same in terms. On liability: "In no event shall our total liability exceed the amount you have paid us in the last 12 months". A company licensing its records to AfterQuery has paid it nothing, so under those words its liability to you for a breach that exposes your records would be capped at nothing. Our reading: the website terms were not written for a company licensing its records, but a company that applies through the website is a user of it. So ask that the data agreement state that it alone governs your records, that it replaces the website terms where they conflict, and that it sets a liability cap of its own.

What to ask for in the licence follows from what the market prices:

• Term. Disclosed AI licences run two to three years or at least three years, and licensing counsel recommends pricing a perpetual right to models already trained separately from a finite licence to the data.
• Exclusivity. On the market's evidence an exclusive licence fetches 2.0 to 5.8 times a non-exclusive one. AfterQuery publishes nothing on it; if it asks for exclusivity, that is a separate price.
• Onward use. Whether labs receive the records or only what AfterQuery builds from them, and a promise that recipients will not attempt re-identification, will not resell the raw data, and are bound by the same terms. Where the records hold bulk US sensitive personal data, ask too that none of it reaches a country of concern under the Justice Department's data security rule, which applies whether or not the data is de-identified.
• Earlier and later licences. Whether the agreement leaves you free to license the same or other records elsewhere. Another buyer's terms may ask you to promise you have not already licensed the same data on similar terms, which matters if you license to AfterQuery first.
• The end. What is deleted, what survives, and by when.

The licence terms page, the guide to data licensing agreements and the guide to exclusive and non-exclusive licences explain each one, and our offer check reads a written offer against AfterQuery's published clauses.

Applying directly or being introduced

There are two ways to reach AfterQuery's data partnerships team. Both lead to the same NDA, the same sample review and an agreement you sign with AfterQuery.

Applying directly

AfterQuery's application form, "Monetize your data", is four short steps: your company (name, website and size), who it should talk to (it sends the mutual NDA to that address), your data, and a review; it says it takes about two minutes and asks for nothing confidential ( its application form). Before you send it, have four things ready: a list of the systems you use and how far back each goes, your headcount now and at your largest, a short note of what you will not share (named clients, health records, privileged files, a director's mailbox), and who in the company can sign. Use an address at the company's own domain. A company that applies itself saves the buyer whatever it sets aside for an introducer, and buyers budget a share of a deal, or a fixed amount for whoever brings them a company, so a holder that comes direct can ask for that budget in its price, though nothing AfterQuery publishes says it agrees to that.

Being introduced

AfterQuery runs a referral programme open to anyone, through a form with no account: "Know a business whose teams work across connected systems? Tell us who they are and how to reach them. We'll handle the rest" ( its referral form). The form asks the introducer for the company's name, website and contact, how they know it and why it could be a good fit, and says: "We will notify you via email if your referral makes sense to learn more about. We track all referrals so payment is ensured if we move forward with the partnership." No window, first-to-introduce or duplicate rule is published. The company pays nothing for being introduced and signs its own agreement with AfterQuery.

This is how an introduction through us works. Value your data first: if AfterQuery's rules take your company, the valuation says so and offers one button, Introduce me. Pressing it is your consent to the introductions it names, and we file each one in the partner's own referral form within one business day. AfterQuery then writes to you itself, and from there it deals with you directly: the NDA, the questionnaire, the sample, the terms and the payment are between you and AfterQuery. The same press introduces you to every other partner whose rules take your data, which is the practical difference from applying to one buyer: several answers to compare, not one. For what each route costs, read going direct or through us.

AfterQuery beside Mercor, micro1, Handshake AI, Turing and Scale AI

A company weighing AfterQuery usually weighs the other large suppliers of human data to labs that buy company records. Each has a guide of its own: Mercor, micro1, Handshake AI and Turing's Project Lazarus.

Turing is AfterQuery's nearest match on who it takes: it also welcomes operating, winding-down and closed companies and also prices from a sample, reviewed within 5 business days. It pays in a different way: it licenses your record to labs and pays a share of every licence sold, nothing upfront, and for the first 90 days only Turing sells it. A company choosing between them is choosing between a price AfterQuery will name after its sample and a share of sales Turing cannot promise.

Mercor is the opposite of AfterQuery on publishing price: its listings put a range on each kind of company, so a company can see before it applies whether Mercor's numbers are worth its while. It is as silent as AfterQuery on the licence, and faster and clearer on payment. micro1 publishes tiers and the broadest licence of the group: perpetual, irrevocable and paid in full once, with the right to sublicense through multiple tiers. It asks for 30+ full-time employees and English-language records, a bar many companies AfterQuery welcomes do not reach. Handshake AI is the one that binds itself in writing to training and evaluation only, and the most specific on how payment is scheduled. Scale AI is the one to compare on handling: it anonymises inside your systems before extraction, where AfterQuery scrubs after transfer.

Read side by side, AfterQuery is the most open of the large buyers about which companies it takes and among the least specific about everything after the sample: price, licence and payment. That is not a reason to avoid it. It is a reason to get its offer in writing beside at least one other, so that its number, its licence and its payment terms can be compared with something. The guide comparing every company-data buyer programme covers a dozen more buyers, the buyer board shows each one's published rules, and the buyer match tool checks your company against them.

Worked example: a freight brokerage past its peak

Example. A freight brokerage in Memphis, Tennessee, has 18 people today and had 45 at its peak, before the freight downturn. It has nine years of loads in its transportation management system, quotes and carrier email in Microsoft 365, a Slack workspace its brokers used for every load, HubSpot for its shippers, and QuickBooks for invoicing. Its owner reads that AfterQuery licenses logistics records and wants to know whether the brokerage fits, what it might earn and what it must settle first.

Does it fit? AfterQuery names "Logistics and field services: dispatch, routing, work orders, customer communication", and asks for several teams working across several systems, and a few years of history. On our reading the brokerage is counted at its largest, the 45 people whose teams made the records, and it has nine years, so it fits. On today's staff it falls below Handshake AI's 20+ full time employees, micro1's 30+ full-time employees and Scale AI's preferred 40+. Mercor's data page names logistics among its industries ( shipping manifests, dispatch logs, invoices) but has no listing that prices a brokerage, and Turing names no size rule. So AfterQuery and Turing are the buyers whose published words most clearly take it.

What might it earn? AfterQuery will not say until it has seen a sample. What raises the figure, in its own factors, is what the brokerage has most of: nine years of history, five connected systems, and a kind of work that is hard to recreate, the negotiation of each load from a shipper's quote request through carrier calls and rate confirmations to delivery and invoice. A load that can be followed from the TMS to the Slack thread to the carrier's email to the invoice is the linked record buyers pay for. On the form's question "What do you have in mind?", the owner answers "Not sure yet" and gets an independent valuation first, so that the first number he names rests on something.

What must it settle first? Its records are full of other businesses and other people. Shippers' names and rates are customer confidences; carriers' rate confirmations are its counterparties' terms; drivers' names, phone numbers and locations are personal data. AfterQuery says customer identities are masked or excluded, but the brokerage decides first which shippers' contracts forbid sharing, whether any carrier agreement does, and what its staff are told. It checks its software terms for Microsoft 365, Slack, HubSpot and QuickBooks, and reads its TMS contract for who owns the data held in it. Payroll it leaves out, as AfterQuery says it never uses payroll anyway. Our page for freight brokerages and 3PLs sets out the rest.

What does it ask AfterQuery? For the NDA before the questionnaire, and whether the sample counts as shared; where the identifiable copy is held after transfer and when it is deleted, and the re-identification test result; the licence's term, exclusivity and onward use, and whether labs receive the records or only what AfterQuery builds; when it is paid and on what event; and that the data agreement replaces the website terms' liability cap. It puts the same records before Turing, whose revenue share answers the same question differently. It may sign with AfterQuery. It will know what it signed.

Questions people ask

What is AfterQuery?

A San Francisco AI training-data company, founded in 2025, that sells expert-built training data, RL environments and custom evaluations to frontier AI labs and enterprises. It raised a $30 million Series A in April 2026 and was reported in September 2026 at a $3.2 billion valuation. Its data partnerships license companies' records of how their teams work.

Does AfterQuery buy company data?

Yes, as a licence rather than a purchase of ownership. It licenses records of how your teams work, from email and chat to CRM and finance, to train and evaluate frontier AI, and its website terms say data you provide remains your property, subject to the licences it needs. It takes operating, winding-down and closed companies.

How much does AfterQuery pay for company data?

It publishes no figure. It prices by depth, breadth and uniqueness: years of history, number of connected systems, the kind of work captured, and how hard it would be to recreate, and proposes terms after reviewing a sample. Its application form asks for the price you expect, in bands from under $50k to more than $500k. Get an independent valuation before you answer.

Does my company qualify for an AfterQuery data partnership?

AfterQuery says it is "most interested in established companies with several teams working across several systems, and a few years of history". We read that as at least 10 people at the company's largest and at least 3 years; the numbers are our reading, not AfterQuery's, and its own team decides at review. It names no country, language or industry it excludes.

Does AfterQuery take closed companies?

Yes: "Operating, winding-down and closed companies are all welcome." A closed company needs someone with the right to sign for it, which may be a former officer still winding it up, a trustee or a liquidator; in Delaware a dissolved corporation continues for three years to wind up. Read who owns a dissolved company's data first.

What does AfterQuery do with personal data?

It says names, contact details and other identifiers are removed or replaced with consistent pseudonyms, customer identities are masked or excluded, and payroll, privileged legal material and credentials are never used, with a re-identification test on every dataset before delivery. The removal happens after transfer, on AfterQuery's side, and only the team working on your partnership sees the data, under NDA. Your own duties to staff and customers stay with you.

How long does AfterQuery take, and when does it pay?

From first form to signed agreement is usually a few weeks, depending on how fast a sample is shared and reviewed, and it replies to a new form within three business days. Payment is the last of its six steps, after transfer and de-identification, with no published schedule, so put the dates and the trigger in the agreement.

Is AfterQuery legitimate?

It is an established, venture-backed company: its Series A was led by Altos Ventures, and TechCrunch reported its later round. Its data partnerships programme is new, and no company has yet confirmed in public a licence of its records to it. Sign the mutual NDA before you share anything, ask which company signs the agreement, and read the agreement itself rather than the website.

How do AfterQuery referrals work?

Anyone can refer a company through its referral form, with no account. AfterQuery contacts the company itself, tells the introducer only whether the referral makes sense to it, and says it tracks referrals so payment is ensured "if we move forward with the partnership". The company pays nothing for being introduced and signs its own agreement with AfterQuery.

What this means for you

AfterQuery takes more kinds of company than almost any large buyer, including small, shrinking and closed ones, and publishes less than most about what it pays and what it takes in return. That makes it easy to qualify and hard to judge from its pages alone. Prepare for the second before you answer the first.

1. Check the fit honestly. Several teams, several systems, a few years: count your people at your largest, list your systems and how far back each goes, and see which of AfterQuery's nine industries your work falls in. The buyer match tool runs the same check across every buyer's published rules.
2. Get your own number first. AfterQuery's form asks for your expected price before it names one. Value your data against published prices and sales, and answer from that, or say "Not sure yet".
3. Decide the scope before the sample. Which teams, which systems, which years, and which records inside them stay out; settle what a later scope would be paid. Use the readiness check and the pages for your systems under rights.
4. Ask the handling questions. Where the identifiable copy is held after transfer and when it is deleted, and the re-identification test's result; our de-identification standard is the checklist.
5. Get the licence and the payment in writing. Term, exclusivity, onward use, what survives the end, when each payment is due and on what event, and a liability cap that replaces the website terms'. Put the written offer through the offer check and decide where to stand with the negotiation page.
6. Hear from more than one buyer. Apply to AfterQuery directly, or be introduced to it and every other buyer whose rules take your data with one press after your valuation. Compare price, licence, de-identification and payment together, as the guide to every company-data buyer programme and the buyer board do.