Frequently asked questions

Maelstrom AI is a professional matching platform that skips the top of the funnel. Instead of browsing a feed and hoping to be noticed, you state an intent — your position and the counterparty you're seeking — and Maelstrom introduces you only when it finds someone whose intent points back at yours. Both sides must accept before any conversation opens.

LinkedIn is a broadcast medium: everyone performs visibility and hopes the right person notices. Maelstrom is a matching engine: nobody browses, nobody gets browsed, and every introduction is reciprocal by construction. There is no feed, no cold outreach, and no way to pay for access to someone's attention.

One hundred and forty-eight positions across twelve categories, each built from three choices: what you are, who you’re looking for, and what for. Capital and investing (venture capital, private equity and growth funds, family offices, angels, institutional LPs, private credit, secondaries, accelerators, placement agents, M&A advisers and operating executives, and the startups and businesses they back), M&A (sellers and acquirers), supply chain (brands, manufacturers, suppliers, buyers), distribution (brands, retailers, distributors), partnerships, media (advertisers, publishers, sponsors, creators), talent (employers and professionals), services, marketing (agencies and the businesses that hire them), licensing, franchising, and mentorship. Every position has exactly one complement, and matching only ever happens across that pair.

No. It's built for any business position. A franchise operator looking for a franchisor, a supplier looking for buyers, an employer looking for candidates, and a publisher looking for advertisers are all first-class users. Startups and investors are simply the pairing people recognize fastest.

Hiring platforms cover talent. Affiliate networks cover media. Investor databases cover capital. Each solves one pairing. Maelstrom covers the pairs — all of them — in one place, on one model.

Yes, across your counterparts, messages, intents, meetings and private notes. It's a proper index rather than a scan, so it stays instant as your history grows, and it understands word forms — searching "commit" finds "committing".

Never. It only reaches your own — there's deliberately no search of the wider network, because that would let anyone enumerate who's here, which is exactly what the double opt-in prevents.

Yes — the box at the top of this page filters as you type, and matching answers open automatically. Press / anywhere on the page to jump to it, and you can link straight to a search with ?q=refund.

Yes, and each gets its own workspace — its own introductions, counterparts and meetings — so raising a round doesn't get tangled up with finding a manufacturer. It's a filter over the same data, not a separate account.

No. Your weekly allowance is shared across everything you're working on, not multiplied. If one is soaking up introductions you'd rather spend elsewhere, pause it.

Yes — a portal for PE firms, VC funds, family offices and private credit, with positions for deal sourcing, co-investment, LP fundraising, secondaries and private credit. Investors can also match with each other rather than only with companies.

Yes. Co-investment is a position that matches with itself, and there are complementary positions for fund managers raising from LPs, and for buying and selling fund positions in secondaries.

Yes. Pick your position on the home page and we'll tell you how many people are seeking exactly what you offer — no account, no email, nothing stored. Counts are rounded into bands and small groups are hidden, so the answer never identifies anyone.

Because an exact number leaks information. If a narrow search returned "1", that one person could be identified by anyone probing the same filters. Bands with a minimum group size prevent that, at the cost of precision we don't think you need.

Yes. If the check comes back with too few people to report, you can leave an email address and we'll write to you once — the first time someone holding the complementary position states a matching intent. No account, no newsletter, and the record is deleted after six months if nobody appears. Every email has a one-click stop link that needs no sign-in.

We say so. You'll see that too few people match to report, and why — never a zero, and never an invented figure. If we haven't measured something, we won't show you a number for it.

You state an intent through a short conversation with Maelstrom AI, which captures your position, what you're seeking, and your criteria — industry, region, deal size, stage, timeline, dealbreakers. Matching runs only between complementary positions and scores each pairing on how well the two sets of criteria overlap. When a pairing clears the threshold, both sides see why, and a conversation opens only if both accept.

An intent is a directional statement: I am X, seeking Y, under conditions Z. It's the atomic unit of the platform. One person can hold several at once — a founder can be raising capital, looking for a manufacturer, and open to acquisition conversations simultaneously, each matched independently.

No conversation opens unless both people accept the introduction. If you decline, the other person is never told — declines are silent. This is why nobody can cold-message you on Maelstrom and why every open conversation is one both parties wanted.

Every introduction shows both sides the exact criteria it was made on, cited back in your own words — shared industry, overlapping deal size, matching region. It's not a black-box "people you may know." It also helps you refine your intent when a match misses.

Maelstrom tells you honestly and keeps watching. You'll be notified when a complementary counterparty joins. Waiting with a reason is better than an empty feed that pretends to be busy.

Time-to-first-meeting is the metric the whole platform optimizes — not messages sent or connections made. With calendars connected on both sides, Maelstrom proposes times that are already free for both of you, which removes most of the delay between matching and meeting.

A room is a private, invite-only matching pool. You create one, share a single link or code, and can then run a room-exclusive search that matches only against people inside that room — ignoring the open network entirely.

Accelerator and cohort programs, conference and trade-show floors, VC portfolios matching their own companies to each other, trade associations offering member-only matching, M&A advisors curating buyers and sellers, and franchise networks matching franchisors with vetted operators.

No. Joining a room is free on every plan. Only creating rooms requires Pro or Team. If invitees needed a paid plan to accept, the invite link would be useless — so it's deliberately open.

No. Rooms use the same complementary-pair rule and the same scoring as the open network. A room narrows who is eligible, never how pairing works. Room-sourced introductions carry an extra tag naming the room, and still require double opt-in.

Yes — room searches draw on the same weekly allowance. If a room surfaces more matches than your plan allows, the extras are reported rather than silently dropped, so you always know what you're not seeing.

Everyone you've matched with, in one place. Each one shows their profile, your conversation, and your meetings — and you can switch between people without losing your place. It's deliberately not called "connections": collecting connections is the behaviour Maelstrom exists to replace. A counterpart is someone who occupies the corresponding position to yours, which is the only reason you met.

Yes — that's what Counterparts is for. The conversation stays open indefinitely, and you can schedule as many further meetings as you like. A match isn't a one-off introduction; it's a working relationship you can return to.

Yes. Each counterpart shows the intent pairing that produced the match — what you were seeking and what they were — so you never have to remember why you know someone.

The relationship moves through warmth states you can see — active, cooling after two weeks, dormant after three. You get a nudge before anything happens, and after 30 days without contact it moves to your archive. Archiving is not deletion: the whole conversation is kept and one click brings it back.

So the list stays meaningful. A contacts list that only grows becomes a graveyard of people you never speak to, and then you stop looking at it. Archiving keeps your active list to the relationships that are actually live.

Not if you mark it as one to keep. Some business relationships are genuinely patient: "come back after your next raise" is a normal thing for an investor to say, and an M&A conversation can pause for a quarter. Marking a counterpart as kept means it never auto-archives, however long it stays quiet. Anything with a meeting already booked is also protected automatically.

Yes, in one click, with the entire conversation intact. Nothing is deleted when a relationship is archived — it just leaves your active list.

Yes, any time, from their Counterpart view. It uses the same mutual-availability engine, so it only proposes times free for both of you and shows both timezones.

Yes. You can state intents from your CRM, act on introductions from your automation platform, read counterparts, propose meetings, run room searches, and receive events as webhooks. The full reference is public, with an OpenAPI spec you can point a client generator at.

There's a free monthly allowance on every plan, including Free — 1,000 units a month, rising to 25,000 on Pro and 100,000 on Team. Usage is metered in units rather than calls, because a matching run costs far more to serve than reading a profile.

No. When the allowance runs out, requests stop with a 402 — they don't silently continue at a charge. To go beyond the allowance you must both add a payment method and explicitly switch on metered usage, and you can set your own monthly ceiling on top of that. We also email once at 80%.

$0.50 per 1,000 units once you've enabled it. Deliberately cheap — the point of metered usage is to avoid a wall in the middle of your workflow, not to be a revenue line.

Create one in the app under Developer API. Keys are scoped, so a read-only integration can't write, and the secret is shown exactly once — we store a hash, so nobody can recover it, including us. Revoking takes effect immediately.

Yes — register an HTTPS endpoint and choose which events to receive: introductions offered, matches created, messages, meetings confirmed or cancelled, relationships going quiet. Deliveries are signed with HMAC-SHA256 so you can verify they came from us. Failures retry with backoff.

Through OAuth rather than an API key. A key is you acting on your own data; an app installed by many people needs each of them to consent individually. Register an app to get a client ID and secret, send users through the authorization flow, and exchange the code for tokens that act as them, limited to what they approved. PKCE is mandatory and refresh tokens rotate. Full details in the OAuth section.

Only what you approved when you connected it, and never more than you could do yourself. Your dashboard lists every connected app with its permissions in plain language, and flags the ones that let it act as you — sending messages, accepting introductions, booking meetings. Revoking takes effect immediately and affects only your account.

Your dashboard runs the same checklist a reviewer does, before you submit — description, website, tagline, privacy policy, HTTPS-only redirects, and whether the permissions you request are defensible. None of it is secret, so you can clear it in advance rather than waiting for a rejection.

Yes. Developers can create a sandbox: ten synthetic people arranged in complementary pairs, with real matches, conversations, and meetings between them. Test keys act as one of those people and can never touch a live account.

The app directory. Every listing shows what an app can access and how many people have installed it before you connect it — so you can judge it without granting anything first.

No. A connected app is capped at half your account's rate limit and half your monthly units. If an app misbehaves it hits its own ceiling and stops — your own keys and your other integrations keep working, and the error names the app so you know which one to revoke.

It can, but never silently. Any additional permission means going through the approval screen again — and that screen leads with what's new rather than re-listing what you already approved, so a widened permission can't slip past you in a list you've seen before.

No, in two independent ways. An app can never use a permission you did not approve — your grant holds exactly what you agreed to, and a developer changing their app cannot reach into it. And if an approved app asks for more than review approved, it is removed from the directory and sent back for review rather than keeping its reviewed status.

Apps start in development mode, usable only by the developer who made them. Going public requires review. If you connect an app that hasn't been reviewed yet, the consent screen tells you so before you approve anything.

Yes, by design. Endpoints are declared alongside their scope and cost in one registry, and both the documentation and the OpenAPI spec are generated from that registry — so a new feature gets a documented, scoped, metered endpoint as part of building it, not as a follow-up someone might forget.

Free is $0 forever: one active intent and three introductions a week. Pro is $29/month: five concurrent intents, fifteen introductions a week, priority in the match queue, anonymous intents, private rooms, and profile customization. Team is $79 per seat per month: everything in Pro for each member, plus a verified organization, match routing, a pipeline dashboard, and pooled quotas.

Yes, and it's genuinely usable rather than a teaser — one active intent, three introductions a week, full match reasons, private conversations, scheduling links, and the ability to join rooms. Paid plans buy capacity, not access.

Capacity and customization: more concurrent intents, more weekly introductions, priority in the match queue, anonymous intents, the ability to create rooms, profile banners and colors, and organization features. We never sell access to people — there is no pay-to-message on Maelstrom.

Scarcity is quality control. If anyone could blast the network, introductions would become spam and nobody would open them. Limiting them means each one is considered — which is what makes an introduction from Maelstrom worth reading.

Paddle.com Market Limited is our merchant of record. Paddle sells the subscription, processes payment, issues your invoice, and handles sales tax and VAT. We never see or store your card details.

We approve every refund request made within 14 days of a charge, no reason required. Cancel any time; access continues to the end of the period you've paid for. Full details are on the Refund Policy page.

Team is billed per seat. Each member gets a complete account inside the organization — their own intents, matches, calendar, and conversations. Introduction quotas pool across the organization, so a heavy month for one person can draw on unused capacity elsewhere.

A company creates one central account and invites members by email. Each member runs a full account inside it — own login, own intents, own matches, own calendar — while the organization is treated as a single verified entity. The company's verification and identity cascade to every member.

Yes, in detail. Every capability is an independent toggle: editing their own profile, changing their own picture, creating intents, accepting introductions, using anonymous intents, creating rooms, sharing room links, connecting a calendar, receiving routed matches, routing matches to others, viewing the pipeline, managing members, changing permissions, editing the org profile, and managing billing. Presets (Owner, Admin, Member, Scout, Interviewer, Managed, Viewer) are starting points you can customize freely.

Yes. Organizations can set profiles to be fully managed, in which case an administrator assigns each member's display name, picture, title, and bio rather than the member doing it. It can be set org-wide as a default or toggled per member. Everything else — their matches, conversations, and calendar — still belongs to them.

Yes — this is what match routing is for. A scout can search for and screen matches, then route them to a colleague's calendar, so the person who takes the meeting is the interviewer, closer, or specialist rather than the person who did the searching. Routing is configurable per member.

Yes. Every member connects their own calendar and matches use their availability by default. If a member's matches are routed to someone else, scheduling uses the destination person's calendar instead.

Pipeline metadata across the organization — how many intents each member has active, how many introductions are pending, how many matches formed, how many meetings were booked. Administrators cannot read the contents of private message threads. It's pipeline visibility, not surveillance.

Their intents are paused rather than deleted, their matches and conversations stay with them, any routing pointing at them is cleared, and their account reverts to the free tier. Nothing is silently destroyed.

Yes. The owner can transfer ownership to another member, who receives full control including billing; the previous owner becomes an administrator. There is always exactly one owner.

Everyone can set a name, headline, bio, location, website, and profile picture. Pro and Team accounts can additionally customize a profile banner and accent color. Organizations can set a company logo, banner, and colors that carry across the company profile.

Members of an organization appear on the company's public profile with their name, headline, and title, so a visitor sees the team as part of the organization rather than as unconnected individuals.

Yes. Individuals and organizations can switch to a private profile, which replaces the page with a "this profile is private" screen for anyone who isn't you or, for organizations, isn't a member. Your connections and details stay hidden. Matching still works — privacy governs who can browse you, not whether you can be matched.

Not unless you allow it. Connections are never a public directory, and a private profile hides them entirely.

On paid plans you can cloak an intent so it's described but not attributed — your name and headline stay hidden until both sides accept. It exists for sensitive situations: quietly exploring a sale, job-hunting while employed, or changing distributors. Note that the description text is still shown, so keep identifying details out of it while cloaked.

That your networks overlap — someone has vouched for you both, or you've both previously matched with the same person. We never say who, and it is not a reference: it tells you nothing about that person's conduct or their business, only that you aren't complete strangers.

No, and we withhold the signal specifically to prevent that. If your own network is small enough that "one person in common" would identify a specific individual, we don't show it at all. The signal appears only when it can't be traced back to someone.

It nudges ranking slightly and is capped, so it can break a tie between two similar counterparts but can never lift a poor fit above a good one. Fit comes first.

Four tiers. Tier 0 confirms you control an inbox. Tier 1 means your email is on a company domain rather than a consumer or disposable mailbox. Tier 2 means you proved control of that domain, either with a DNS record or by signing in with a Google Workspace or Microsoft 365 account on the same domain. Tier 3 means a person reviewed your business registration details. Separately, government ID verification confirms you're a real individual.

Control of an inbox, a domain, or — at the highest tier — that submitted registration details checked out. A badge is not a background check, a credit check, or an assessment of anyone's honesty, solvency, or authority to transact. Someone with the highest badge can still be a poor counterparty, and domains change hands. Always do your own due diligence before moving money or sharing confidential material.

Yes, through a specialist verification provider. The provider performs the check — document, liveness, and face match — and returns only a result. Maelstrom never receives or stores your document, ID number, or date of birth. Verified identity earns a badge, unlocks interaction, and enforces one account per person.

No. Verification is free for you, and it's free for us too within our provider's monthly allowance — so there's no incentive for us to ration it or push you toward a paid plan to get verified.

Three layers: email confirmation, work-domain and domain-control checks, and — for identity-verified accounts — a one-person-one-account rule enforced through a salted hash of the verification result. We can tell that the same verified person has signed up twice without ever storing anything identifying about them.

Slightly. Higher tiers get a small ranking bonus, deliberately capped so that a badge can reorder results but can never manufacture a match your stated criteria don't support. If verification could create matches, buying a domain would become an attack.

No, and we say so plainly. We don't confirm that someone holds the position they claim, controls the capital they describe, or intends to deal honestly. Double opt-in means nobody can cold-message you, but the due diligence before a deal is yours. Be especially wary of pressure toward speed, secrecy, upfront fees, or moving off-platform early.

When you pick your position, we show you a worked example from someone in the same one — a real headline, a real summary with numbers in it, and the criteria that go with it. It's there to show the shape, and you can hide it.

Whatever you've typed is saved on your device as you go, and it's still there when you come back. You'll also see a bar telling you you're offline rather than a button that silently fails.

Write one sentence about what you're seeking. That's it. There's no profile to fill in and no network to build first — the dashboard walks you through the few steps that matter, in order, and gets out of the way once you're going.

No. It shows the single most useful next step, disappears once you've done everything, and can be hidden at any time — after which it stays hidden.

Yes. You can sign up, state an intent, see who you match with and exactly why, edit your profile, connect a calendar, and set your availability — all before verifying anything. Verification is required only when you want to reach another person: accepting an introduction, sending a message, joining a room, or booking a meeting.

Because the value of an introduction here is that it came from a real, accountable person. Anyone can explore; nobody unverified can consume your attention. It also enforces one account per person, which is what stops someone flooding the platform with fake profiles.

First, confirm your email — that activates introductions and lets you create rooms. Second, verify your identity with a government ID through our provider — that unlocks accepting introductions, messaging, joining rooms, and booking meetings. Your dashboard shows exactly which step you're on and what it unlocks.

Yes. Declining is always available — only accepting reaches another person. Declines stay silent either way.

Yes — Google Calendar or Microsoft 365. Once connected, Maelstrom proposes meeting times that are already free on both sides' calendars, so matches turn into meetings without scheduling back-and-forth.

It depends on what you connect. Google Calendar: free/busy only for finding times, so we never see your events' titles, attendees, locations, or descriptions, plus the events Maelstrom itself adds for meetings you confirm. Microsoft Outlook: Microsoft offers no free/busy-only permission, so this connection could read your other events; Maelstrom keeps only start and end times and opens only the events it creates, and the connect card says so before you click. Zoom: only the meetings Maelstrom creates. Availability from a connected Google Calendar is computed when you ask and isn't retained. There is one exception, and we'd rather state it than let you find it: if you subscribe to a published calendar URL instead of connecting Google, we cache the busy intervals from that feed — start and end times, nothing else, never titles or attendees. Fetching a third-party feed inside every request would be slow and would hammer someone else's server. The cache refreshes about hourly, is deleted the moment you remove the feed, and goes with your account if you delete it.

Only the specific times you choose to propose. They never see your calendar, your busy blocks, or anything you didn't send.

Yes. You can add a Cal.com, Calendly, or Zoom link to your account, and it appears in your match threads so counterparties can book you directly — no calendar connection required.

Every proposed time is shown in both people's timezones, side by side — because "2 PM" isn't a time until you say whose. We use real named timezones (like America/New_York), not fixed offsets, so daylight saving is handled automatically: your 9am stays 9am when the clocks change, on either side of the world.

You'll see a warning saying so explicitly. Maelstrom also warns when a time falls outside normal working hours for either side, when you're more than eight hours apart, and when the other party's zone has a partial-hour offset like India's +5:30 — the cases people get wrong doing the maths in their head.

Yes. Set your timezone, your available hours for each day of the week, your default meeting length, a buffer around existing events, minimum notice, how far ahead people can book, and how many proposals may land on any one day. You can also block specific dates like holidays.

Maelstrom tells you plainly that no time works inside both sets of stated hours, and — when you're many hours apart — says that widening your hours is what usually unlocks it. It won't invent a time that doesn't suit one of you.

Yes. Any confirmed meeting can be rescheduled or cancelled. Rescheduling proposes a new time and asks the other side to confirm again — they agreed to a specific slot, not to whatever it becomes. Cancelling sends a calendar cancellation so the entry disappears from their calendar automatically.

Yes. You can set how much notice someone must give you before moving a meeting, and how many times a meeting may be moved at all. Both have sensible defaults — 24 hours' notice and two changes — so you don't have to make a decision about it before you've ever had a meeting. Your settings protect you: it's the other person's policy that applies when you try to move a meeting on them.

No. The notice required is capped by the lead time the meeting was booked with, so a meeting arranged for tomorrow morning can still be changed even if the other person normally asks for 48 hours. A rule that's unsatisfiable from the moment it's created would be a trap, not a policy.

No. Every confirmed meeting comes with a calendar file you can add in one click, which works with Apple Calendar, Outlook, Google Calendar, or anything else. Connecting a calendar is optional and only adds automatic availability checking.

Only if you opt in. By default we request free/busy access alone, which cannot create or change anything. If you tick the box for event writing when connecting, we'll add confirmed meetings for you; if you don't, you get the calendar file instead. Neither option lets us read your event contents.

When times are shown to you, they're held for ten minutes so they aren't simultaneously offered in another conversation. If you don't use them the hold expires by itself and the times return to circulation.

If you connect a Google Workspace or Microsoft 365 account whose domain matches your email domain, it verifies you to Tier 2 automatically — your identity provider vouching for the employment relationship, which is harder to fake than a DNS record.

You invite the counterpart you're missing — a founder invites an investor, a brand invites a manufacturer. The invite is tied to a specific intent, so the person receiving it sees exactly what you're seeking rather than a generic "join my network" message. They can decline in one click without joining.

Because it wouldn't help you. On a matching platform, inviting people like yourself thickens one side of the market and makes matching worse for everyone, including you. The useful invite is the one that fills the gap you're facing.

Three a week on Free, ten on Pro, twenty-five on Team. The cap is deliberately low — invites are a matching tool, not a growth channel, and a generous limit is an invitation to bulk-mail.

An introduction credit — but only once they state an intent, never merely for signing up. Rewarding signups rewards volume, and that's how referral schemes turn into spam. It also has to be the position you were actually short of.

Nothing yet. The invite shows you what that person is seeking and what position they think you hold. If you join and state an intent, you'll be considered — and you'd both be asked separately before anything is shared. Joining does not introduce you to them automatically.

Click decline on the invitation. It's permanent and platform-wide: nobody on Maelstrom will be able to invite you again, not just the person who tried. The sender is told only that it was declined, never why.

Yes. Each kind — introductions, messages, meetings, reminders, product news — can be set to instant, a daily digest, a weekly digest, or off. You can also set quiet hours; anything that arrives during them waits rather than being dropped.

For the things that can't be undone — deleting your account, removing an org member, suspending an app — we schedule rather than do. You get an email straight away with a one-click link to stop it, and you don't need to sign in to use it. Account deletion waits 24 hours.

There are limits on how much anyone can send — overall, per conversation, and in any single minute — so flooding isn't possible even before you report it. If someone is being a problem, report them; we act on it.

Account and security messages — sign-ins, recovery, billing. Someone who had silenced those couldn't be told their account was accessed, which isn't a setting we're willing to offer. Everything else is yours to set.

Pause your account. No new introductions arrive, your intents pause rather than being deleted, and the inactivity clock stops so nothing gets archived while you're away. Set a date and it resumes itself, or resume manually whenever.

They're told you've paused and roughly until when, and that the conversation is untouched. Silence reads as being ignored — this way nobody's left wondering why you stopped replying.

No. Intents pause and resume exactly as they were, conversations and notes stay, and nothing is archived for inactivity while you're paused. A pause that quietly deleted things wouldn't be a pause.

Yes — PDFs and images up to 25MB, attached to a conversation. Not Office documents — those can carry macros and we don't scan for malware. Export as a PDF.

No, and we say so rather than implying otherwise. We check the file type, its size, and that its contents match what it claims to be — an executable renamed to .pdf is refused. But treat a file from someone you've just met with the same care you'd give an email attachment.

You can remove it from the conversation and it's deleted from our storage. If they already downloaded it, we can't take that copy back — and we won't pretend otherwise.

Not silently. After 90 days we ask whether an intent is still current — one click either way. If we hear nothing, it pauses at 120 days rather than continuing to generate introductions. Nothing is deleted: the intent, its criteria and its history all stay, and confirming it resumes matching immediately.

Because an intent that has moved on costs somebody else too. If you closed your round six months ago and we keep introducing investors to you, that uses their weekly allowance on a conversation about a deal that's finished — and they blame us, correctly. Asking once is cheaper than wasting both of you.

Slightly, before it's ever paused. An intent nobody has touched in months is less likely to be live than one edited last week. The effect is small and floored — editing or confirming it resets the clock entirely.

No. At most one summary email a day, and only when something is genuinely waiting on you — a relationship about to be archived, a meeting needing confirmation, or a saved search that matched. Never one email per event.

Nothing punitive. After 60 days away your intents pause, so you aren't consuming other people's weekly introduction allowance while nobody is home. The moment you sign back in they resume automatically — you don't have to do anything.

It affects your ranking, not your eligibility. Someone who hasn't visited in months is shown lower than an equally good match who is present, because an introduction to someone who never replies wastes a real person's weekly allowance. The effect is floored — return and you're immediately competitive again.

After a meeting passes, both sides are asked whether it happened. A no-show only counts once it's been reported and the other person hasn't disputed it — one person's account is never enough on its own. Repeated confirmed no-shows lower how highly that person ranks in matching, because the cost falls on the counterparties they stand up.

You get the introduction back. The match that produced that meeting came out of your weekly allowance, so once a no-show is confirmed we return the credit — it sits on top of your normal allowance and doesn't expire. The cost of a no-show should fall on the person who caused it, not the person who kept their word.

They can report it, but it doesn't count until you've had a chance to dispute it. Disputing is one click and prevents it counting at all — and no credit is refunded on a disputed report either. If you did miss it, you can admit it instead, which refunds them immediately rather than making them wait out the window. Confirming a meeting happened credits both people, so the honest path is also the easiest one.

Yes, a day after it happens. It's one tap, and if you say it wasn't worthwhile we ask why from a short list. The platform otherwise only measures whether people meet, which isn't the same as whether meeting was useful.

No. They never see it and are never told you were asked. That's the only way to get an honest answer, and honest answers are the point.

It's aggregated to find our mistakes. The reasons are a fixed list precisely so they can be counted — if lots of people say the size range was wrong, that's a scoring fault we can fix, as opposed to two reasonable people simply not clicking.

Yes, and we show you what. If several of your answers point at the same cause, we propose a specific change — "narrow your range to $250K–1M, based on your last three" — and you can accept or leave it. We never edit your intent without you saying so: it's your description of your business, not ours. Everything we've changed is listed under "what your feedback has changed".

We'll tell you if there's a pattern. If several of your mismatches share a cause — size, stage, industry, geography — you'll get a specific suggestion about what to change in your intent, rather than a score.

Every match conversation has a Report button. Choose what happened — misrepresentation, asking for money up front, phishing, spam, harassment, or pushing to move off-platform — and add detail. The report goes to our trust team with the conversation attached, and the person reported is never told.

A person reviews it and can suspend the account. We keep the conversation as evidence. To be direct about the limits: we can act on the platform, but we cannot recover money — if you have been defrauded, report it to your local authorities too.

A verified user can put their name behind another user. Vouches appear on the profile, along with how many came from identity-verified accounts. Mutual vouches — where two accounts vouch for each other — are counted separately and shown, so a ring of new accounts vouching for one another is visible rather than hidden.

Yes — saved searches. Pick a counterparty position and we'll email you when someone matching arrives, instead of leaving you to check back. Free includes one saved search, Pro five, Team ten.

The site sets a strict Content-Security-Policy: scripts are allowed only by cryptographic hash, so injected code doesn't run even if it reaches the page; forms can't post anywhere but here; and the site can't be embedded in someone else's page. It's a second layer under careful escaping, for the one place escaping might slip.

No. If part of the page fails, the rest keeps working and you'll see a short banner saying so with a reload button. Anything you'd typed is saved on your device, and any change you'd already submitted was either sent or is queued to send.

No. It's held on your device and sent when you reconnect. Each queued change carries a unique key, so if a reply gets lost on a flaky connection the retry can't create a duplicate.

Yes, under account settings — each device, roughly where it connected from, and when it was last active. You can end any one of them, or sign out everywhere while staying signed in on the device you're holding. We deliberately don't collect a device fingerprint; a coarse label is enough to recognise something that isn't you.

There's a daily allowance, larger on paid plans. If you hit it, the message says when it resets and nothing else on Maelstrom is affected. The limit exists so one runaway script can't run up an unbounded bill.

Yes. Press ? anywhere in the app to see them. j and k move through your relationships, Enter opens one, n adds a note, and g then r jumps to Relationships. Shortcuts never fire while you're typing in a field.

Yes — each section has its own address, so you can bookmark your billing tab or send someone a link to it, and refreshing keeps you where you were.

Into five tabs: Now, Relationships, Scheduling, You, and Account. Introductions and anything needing your attention are in Now, which is where you land. A number on a tab means something there is waiting on you.

Yes — across your counterparts, messages, intents, and meetings. It only ever covers your own data; there's no search of the wider network, because that would let anyone enumerate who's here, which is exactly what the double opt-in prevents.

Yes, from account settings — a machine-readable file plus a readable summary. It includes your account, intents, conversations, meetings, and everything else we hold about you. A few things are deliberately withheld and the file says which and why: reports other people filed about you (disclosing them would identify the reporter and make reporting unsafe), and the other side's response to an introduction, since declines are silent by design.

Set up recovery before you need it, under account settings. Two options: recovery codes — eight of them, shown once, which you should store somewhere that isn't your inbox — and a verified second email address. Either gets you back in. Without one, losing your inbox means losing the account.

It has to be confirmed from that inbox before it can be used, we only ever show it masked, and adding one sends a notice to your primary address — so nobody can quietly add a way into your account. Recovery codes are stored hashed; we couldn't tell you yours even if you asked.

There's a limit of five an hour per address. It exists so nobody can flood your inbox with sign-in codes you didn't ask for. Wait a little and request a new one.

Check spam first, and if they're there, marking one as "not spam" fixes future ones. If an email to you hard-bounces or you mark us as spam, we stop sending to that address entirely — contact support and we'll clear it.

Passwordless. Enter your work email and we send a six-digit code that works once and expires in fifteen minutes. Type it into the six boxes on the sign-in screen. The same screen creates your account the first time and signs you back in after that, with everything you saved. Five wrong tries cancel a code, and asking for a new code cancels the old one. There's no password to forget, reuse, or have breached.

Yes — email confirmation activates matching. It's also how we keep throwaway accounts off the platform, which is part of why an introduction here is worth opening.

You'll receive a branded invitation email. Accepting walks you through confirming your email, completing your profile, and verifying your identity. You end up with your own full account inside the organization.

We use HTTPS throughout, passwordless single-use sign-in codes, signed and replay-protected billing webhooks, OAuth with PKCE for calendar connections, and AES-256-GCM encryption of calendar tokens at rest. Server-side enforcement means the browser can never claim a plan or permission it doesn't have.

No. We don't sell personal information, we don't share it for advertising, and there is no advertising on Maelstrom. Full detail is in the Privacy Policy.

Yes, from your dashboard — type your email to confirm. We remove your profile, intents, availability, saved searches, and calendar connection. Your messages are replaced with a placeholder so the other party keeps a coherent thread, and safety records plus the duplicate-prevention hash are retained as described in the Privacy Policy. Organization owners must transfer ownership first.

Three jobs: it interviews you to capture your intent in structured form, it scores potential matches, and it writes the short brief explaining why you and a counterparty were matched. It also answers questions about the platform in the chat widget.

No. Our AI provider processes the text to generate a response and does not use API submissions to train its models. If you'd rather not use AI at all, you can enter your intent through structured form fields instead.

No. Scoring decides what you're shown; you decide what you accept. No introduction happens without explicit acceptance from both sides. If you think automated scoring has affected you unfairly, you can ask for a human to review it.