Privacy
Last updated November 15, 2026
Anyone I Know is operated by Hunchwork Labs LLC, a Texas company ("we" on this page). Where it says Michael, it means Michael Cummings, who runs it for the company. This page says, in plain words, what the app keeps about you, what it doesn't do, and how to take your data with you or delete it. Questions go to hello@anyoneiknow.com.
What we keep
- Your account: the email you sign in with, your name, and your settings (the roles you're looking for, your time zone, when to send the morning email).
- What you add: companies, roles, applications and their steps, interviews, notes, timelines, the people helping you and what they offered, and the names of your resume versions with a link or file name. The resume and cover letter files you upload to Documents are kept, as the next point says. A spreadsheet of applications you import is read in your browser, and only the columns you pick (company, role, date applied, status, link, resume name, recruiter, salary, location and notes, plus the other columns you leave ticked to keep in the note) are saved; the file itself isn't uploaded.
- Contacts you import: LinkedIn, phone, Google and Facebook files are read in your browser, and only these details are saved for each person: name, company, title, email, phone, profile link, location and when you connected (whichever the file includes). The file itself isn't uploaded. For each file, a note of when it came in, which kind it was and how many people it added, so Settings can list what came in.
- If you connect your own AI assistant (Claude, ChatGPT and others): your assistant reads your email on its side, with its own access. It can save to Anyone I Know the job emails it matches to one of your applications (sender, recipient, subject, date and text), and, for each change it suggests, a short excerpt of the email it came from. Nothing else from your inbox is stored, and you can disconnect it at any time. If you set up by chatting with it, each change it saves for you keeps the words you said for it, so you can see where it came from and undo it; they're shown only to you. If your assistant shows your Inbox as a card in the chat, what you tap on it (save, skip, ask, or moving on to the next role) is kept with your account, so the connector can tell that assistant what you did; nothing else from the chat is kept.
- If you turn on email forwarding (Settings → Forward job email): you get a private address, and a Gmail filter forwards only job mail to it (mail from recruiting systems, with job words in the subject, or from senders already matched to your applications). We only read mail your own Gmail forwards to that address. When a forwarded email is about a company you're tracking, we keep a copy on that application (sender, subject, date and text), the sender's address so later mail from it can be matched, and a suggested update with a short excerpt, which changes nothing until you approve it. Mail about none of your companies is dropped without being stored, and the original message is never kept. Matching is done by fixed rules on our server: no AI model reads your forwarded email. If that changes, this page will say so before it applies to you.
- Your resumes and cover letters (Documents): the files you upload (PDF or Word), their words, the PDFs the app makes from them, the copies you make for each job and the new versions written for you, and which one went out with which application. A Word file's words are read in your browser. To read a PDF's words, and for Fix it for me, Tailor for a job and Draft a letter, the app sends the resume or letter to Anthropic (the maker of Claude), whose AI reads it for us, with only what that needs: for tailoring, the job's posting and what you told it there; for a letter, also your notes and answers for that application and the emails you sent about it. Nothing else about you goes with it, and nothing that names you beyond what the document itself says. The PDFs are made on Cloudflare, where the files are kept. We keep a record of each of those AI calls (what kind, when, and what it cost us), so each person's AI use stays under a monthly limit. A version your AI assistant writes with you in a chat is kept only once you approve it. Delete everything removes the files and their words.
- If you start from your resume (setup's search step, or Settings → What you're looking for): the app reads it once, free, to suggest job titles, levels and places for your search. A Word file is read in your browser and only its text is sent; a PDF is sent as the file. Either way it goes to Anthropic (the maker of Claude), whose AI reads it for us, with nothing else about you. Nothing goes into your search until you tick what you want and save. If you use what it suggests, we keep the file and its text in Documents (above), so later features don't have to ask for it again; if you don't, we keep neither, only that you used your free read, when, and how much it cost us. Delete everything removes them.
- Setup: where you are in it, when you said you asked LinkedIn for your connections and when the file came in, so you can pick up where you left off; and, if you ask for them, a time for one reminder email and when we last emailed you the LinkedIn steps.
- If you joined the waitlist (the welcome page had a form for it until October 28, 2026; since then anyone can make an account): your name and email, and whatever you added in the optional second step (your LinkedIn profile link, the roles you're after, a few companies you'd love to work at, where you are in your search, where you're based, who told you about it and anything else you wrote) and your resume, if you added one. Only Michael sees it, and he got an email with your name and address when you joined. When your account is made, it starts from it: it takes your name, setup's first step starts from the roles you're after, your dream list and where you're based (you can change or remove any of it before it's saved), and "Just curious" starts you with the tracker off. Your LinkedIn link and resume stay with the application. To limit spam we also kept a one-way code made from your internet address and the date, never the address itself, and the form ran Cloudflare Turnstile (below). Email hello@anyoneiknow.com to have your application deleted at any time; deleting your account deletes it too.
- If someone's share link brought you (their "Start in Claude" or "Start in ChatGPT" button, or their anyoneiknow.com/start link): that link's code, kept with your account when it's made. To carry it from the link to signing up, the link sets a cookie in your browser for up to a week, holding only the code; it's cleared when you next open anyoneiknow.com signed in (signing up takes you there). The code is random, so it doesn't say who made the link, and Michael sees only how many accounts each code brought. If you scanned a QR code Michael showed you (at a job fair or another event): the same, but the code is his own, one for each event, so he sees that your account came from it. If you make share buttons (Settings → AI assistant), your account gets its own code, which is in every link you share.
- Feedback you send, and simple usage counts: how many times a day you opened each screen, and how many requests from you or your assistant were turned away that day for coming too fast (a limit that keeps one stuck assistant from slowing the app for everyone; kept for 30 days). Feedback you send from the "Something broke" screen, or just after an error, is linked to that error report, and copied (without your name or email) onto the matching fix in Michael's task tracker. Feedback your AI assistant sends for you comes with the last few messages of that chat, which it shows you before sending; they hold whatever that chat said (which can include your contacts or companies), only Michael sees them, and they're never copied anywhere else.
- Page views and load times: Cloudflare, which hosts the app, counts each page and screen you open (in the app, and the public pages: the homepage, sign-in and this one), with its address (nothing after a ?), the site that sent you, your country, browser, operating system and device type, and how long it took to load. Michael sees these as totals in Cloudflare Web Analytics, not as a record of what any one person did. It sets no cookies and stores nothing in your browser.
- Error reports: when something breaks, a report goes to Sentry, a service that collects crash reports. It holds the error and where in the app's code it happened, which screen it was on (the address without anything after a ? or #), your browser and device type, an anonymous account number (not your name or email), and the few steps just before it, like which screen you opened or which button you pressed (its label isn't sent). It never includes your contacts, companies, roles, notes, emails or anything you typed, and sign-in details are removed before it's sent. Reports are kept for 30 days.
What we don't do
- No ads, and we don't sell or share your data.
- No ad trackers, and nothing that follows you around the web. The page-view counts above stay with Cloudflare as part of hosting the app. Apart from Cloudflare's page-view counter, the sign-in page's only outside script is the Turnstile spam check on its form, and the homepage and this page have none.
- Error reports are passed on to Sentry by our server, so Sentry never sees your internet address or location.
- Michael's admin view shows those usage counts for each person, how far each person has got through setup (for example, whether contacts are imported or an assistant is connected, not what they are), the feedback you send him (with the chat messages, if your assistant sent it), and your waitlist application if you joined that way. To see which companies the app can find job boards for, it also lists the company names in everyone's networks put together, each with how many contacts people have there in all, how many people know someone there or have it on their dream list, and its job board, plus, for each person, how many of their companies have a board the app reads. That list never says whose companies they are and never names anyone's contacts. Otherwise it never shows your companies, roles, contacts, notes, emails, resumes or cover letters, except what you choose to send as feedback.
- If you haven't moved on in setup for about a day, Michael and the person helping him with the beta get one short email saying so: your name, which setup steps you've done, when you last used the app, and how many contacts and Inbox roles you have. Never your companies, roles, contacts or anything you typed.
Where it lives and who handles it
- Your data is stored on Cloudflare, in the United States, and a copy on our own server in the US.
- Every night the app saves a copy of its database, also on Cloudflare in the United States, so a mistake can be undone. Each copy is kept for 30 days, or 90 for the one from the first of the month, and then deleted. A copy of each is also kept on our own server in the United States, so the backups don't depend on Cloudflare alone, for the same time, and then deleted too. Your resume and cover letter files are copied the same way, to Cloudflare and to our own server: a file you delete keeps its copy only as long as a database copy from when you had it, so at most 90 days.
- Sign-in is the app's own, below. Cloudflare Access handled it until October 27, 2026 and no longer does.
- The app's own sign-in is in its own database (no outside company holds it). You can sign in with a link and a 6-digit code we email you: we keep the code (encrypted) and a scrambled form of the link for 15 minutes, or until you use one, and, to limit how many sign-in emails one address gets, a one-way code made from your address with the time of each email, for a day. If there's no account for the address, its email says so, and nothing else is kept. While you wait for the email, a cookie remembers for 15 minutes that this browser asked, so the link can sign you in at once here; it doesn't hold your address. The sign-in page's form runs Cloudflare Turnstile, a check that tells people from bots: Cloudflare looks at details your browser sends with it (like your internet address and browser type) only to detect and block bots.
- You can also sign in with Google or LinkedIn. From them the app keeps only your email address, your name (when your account is made, if your waitlist application didn't give one) and that you sign in that way, with their id for you so the next sign-in finds your account. It never keeps your photo or the access Google or LinkedIn hand over: nothing lets the app read your Google or LinkedIn account. It only takes an address they say is verified. While you're at Google or LinkedIn, a cookie and a record of that sign-in are kept for up to 10 minutes so it comes back to the right browser.
- Once you're in, the app keeps a sign-in for each browser you use: a cookie, and a record of when that sign-in started and when it runs out, with no IP address or device. It lasts 30 days after you last used the app, and 90 days at most. The app also remembers a few display choices (like your theme) in your browser.
- The morning email is sent through Cloudflare's email service. You can change its time or turn it off.
- Forwarded job email arrives through Cloudflare Email Routing and is read by the app's own server code on Cloudflare.
- Page-view counts are kept by Cloudflare: in full for 7 days, then as a sample of about one in ten, for up to six months.
- Error reports are stored by Sentry (Functional Software, Inc.) in the United States for 30 days. A summary of each new error (what broke, where in the code, how often), plus any feedback about it without your name or email, goes into Michael's task tracker, Linear (Linear Orbit, Inc.), so it can be fixed.
- The app checks public job boards (like Greenhouse, Lever, Ashby and Workday) for the companies you track and the companies where you know people. To find a company's board, it asks those sites whether one exists under the company's name, and if not, visits the company's own public website to find its careers page. Those checks don't send anything about you. Each company is looked up once for everyone: the app keeps a shared list of the company names it has looked up and the board or public careers page it found, if any, with nothing that says whose contacts or searches they came from.
- To tell where a job is, the app matches its posted location against a list of real places from GeoNames (geonames.org, used under CC BY 4.0), kept in the app's own database. Nothing is sent anywhere to do it.
- If you start from your resume (above), the resume is sent to Anthropic (the maker of Claude) for its AI to read, once, with nothing else about you, and its answer comes back to the app. Your resumes and cover letters in Documents go to Anthropic the same way, each time you ask for one of its AI pieces (above), and for a PDF's words to be read.
- When a company's own careers site isn't on a job board the app already reads, an AI agent can work out once how to read it: the app sends the company's name and its public careers pages to Anthropic (the maker of Claude), which works out the steps, and the app then follows those steps itself every day. To sort companies by industry, the shared list's company names (and the public website found for each, if any) are also sent to Anthropic, which says which industry each one is in; the industry is kept on the shared list. Nothing about you, your contacts or your search is sent; it's the same shared list of company names, with nothing that says whose they are.
Your controls
- Download everything (Settings): a ZIP file with all your data, including stored emails, your resumes and cover letters (the files themselves and their words), and your waitlist application, if you sent one.
- Delete everything (Settings → Your data): removes your contacts, companies, applications, stored emails, notes, your resumes and cover letters (files and words), waitlist application, sign-ins and account, and takes your email off the list of people allowed to sign in. It can't be undone: signing in again would start a new, empty account. Copies of your data in the nightly backups are gone within 90 days, as each copy expires.
- Turn off the morning email, or disconnect your AI assistant, in Settings at any time.
- Turn off email forwarding in Settings at any time, and choose whether to delete the forwarded emails stored on your applications. Remove the forwarding address and filter in Gmail too.
Changes
If what we keep or how we use it changes, we'll update this page and its date, and tell you in the app before it applies to you.