Your drives. One mesh.
Backup that survives failure

Recilic weaves the Macs and drives you already own — at home, at the office,
at a friend's — into one end-to-end encrypted backup pool. Lose a disk,
a machine, even a whole location: your files come back.

Download on the Mac App Store See how it works
FROM THE MAKER OF FOLDIC & RIPPLIC

Encrypted before it leaves

Files are encrypted on your Mac before anything is shared. Nodes hold unreadable ciphertext shards — no keys, no file names, no folder structure. Not even your friends can peek.

Survives failure by math

Erasure coding splits every backup into data + parity shards across your nodes. With 4+2, any four shards rebuild everything — two machines can vanish and you're still whole.

Proof, not promises

"Backed up" means verified: Recilic routinely challenges nodes to prove they still hold your shards, and runs restore drills — so recovery is demonstrated, not assumed.

Bring your own storage

Your spare Mac mini, the office Mac Studio, a NAS, external drives, a trusted friend's machine — idle space you already have becomes your backup pool. No cloud bill, no lock-in.

Location-aware placement

Recilic knows which nodes share a home, a network, a city — and spreads shards across failure domains. Six copies in one house is one fire; a mesh is not.

History that can't be rewritten

Versioned, immutable snapshots with delayed deletion. Ransomware or a slipped finger can't silently destroy your history — roll back to yesterday, or last month.

Three steps, then it takes care of itself

Set up your mesh once. Recilic encrypts, shards, distributes, audits, and repairs from then on.

1

Pick folders

Choose what to protect on your Mac — Photos, projects, documents, everything.

2

Invite nodes

Your other Macs, a NAS, drives at the office, a trusted friend's machine. Each contributes spare space.

3

Rest easy

Every change is encrypted, sharded, and verified across the mesh — and Recilic shows exactly how recoverable you are.

Photos Library
Encrypted locally
6/6 shards verified across 6 nodes
Recoverable with any 4 nodes
Last restore drill: passed 7 days ago
⏱ Latest verified backup: 3 minutes ago

What Recilic proves

"Backed up" is a claim. Recilic backs it with four kinds of proof.

Proof 01

Possession

Recilic routinely challenges every node to answer with the exact bytes of your shards. Each challenge is single-use and bound to one node and one shard — a node that lost your data can't bluff, replay, or borrow an answer.

Proof 02

Restorability

Scheduled restore drills rebuild your files end to end — gather shards, decode, decrypt, verify — proving the whole chain works, including your keys. Recovery is rehearsed, not assumed.

Proof 03

Survivability

Your recoverability report is computed from where shards actually sit: exactly how many machines — or whole locations — you can lose and still restore. Derived from real placement, not wishful defaults.

Proof 04

History integrity

Backup generations are immutable, sequence-numbered, and bound to your vault key. Nothing can silently rewrite yesterday's backup — and nobody can swap your encryption key unnoticed.

Most backup tools' checkmark means "the transfer finished." Recilic's means "your nodes just proved they still hold your data — and a full restore was rehearsed."

How it compares

A NAS protects you from a dead disk. A cloud drive protects you from a dead house — if you trust it with everything. Recilic covers both, with machines you already own.

Recilic meshHome NASLaptop + cloud drive
Fire, theft, disaster Shards spread across locations — lose a whole site, restore anyway Sits next to your Mac; RAID doesn't help when the house does Data lives off-site in a data center
Drive or machine failure Any 2 of 6 shards can vanish; the mesh self-heals RAID survives 1–2 disks, but the NAS box itself is a single point Provider-side redundancy
Ransomware & slipped fingers Immutable history — snapshots can't be rewritten; roll back to any point Snapshots only if configured; a network-mounted NAS often gets encrypted too ~30-day versions; sync faithfully uploads the damage
Privacy Encrypted before leaving your Mac; nodes hold unreadable fragments Yours — but whoever reaches the box reaches the data Provider holds the keys and can read contents
Who you must trust Nobody — below the recovery threshold, nodes reconstruct nothing Everyone with access to the box The provider, entirely
Proof restores work Possession audits + scheduled restore drills Assumed — until the day you need it Assumed — and accounts get locked
Upfront cost $0 — space you already own $300–1000+ for the box and disks $0
Monthly fees None Electricity ~$10/mo for 2 TB — forever, and rising with size
First backup of a big library LAN speed LAN speed Days over home upstream for 1 TB+
Maintenance Self-healing, with an honest recoverability report Firmware patches, RAID scrubs, disk swaps — a hobby Zero
Third-party dependence None — your machines, your keys None Pricing, terms, policies, account bans

Recilic isn't a sync tool or a Time Machine replacement — keep Time Machine for quick local rollbacks. Recilic adds the layer they can't: off-site, end-to-end encrypted, verified disaster recovery on hardware you already own.

For the technically inclined: Recilic vs. Kopia

Kopia answers “how do I snapshot efficiently into a backend I trust.” Recilic answers “how do I get durable, verified storage without a backend to trust.”

RecilicKopia (+ your backend)
Storage model A mesh of machines you own — every chunk erasure-coded 4+2 across nodes and locationsOne repository on a backend you provide (S3, B2, SFTP, a disk); redundancy is the backend’s job
Losing storage Any 2 of 6 shards can vanish; dead nodes are rebuilt from survivors The repo is a single point — lost blobs mean holes in restores; off-site means maintaining a second repo yourself
Trust No single party holds enough fragments to reconstruct — or to hold your data hostage Client-side encrypted, but one place holds 100% of the ciphertext and can delete it
Verification Built-in possession audits (single-use challenges) + automated restore drills kopia snapshot verify — schedule it yourself; verifying means reading data back
Self-healing Rebuilds lost shards from any 4 survivors, automatically Optional ECC guards bit flips; lost blobs can’t be regenerated
Survivability insight Report computed from real placement: how many machines or locations you can lose No equivalent — there is one backend
Dedup & compression Basic today (fixed-size chunks) Excellent — content-defined chunking, compression, cross-machine dedup
AudienceA Mac app for people who don’t want to run infrastructureCross-platform CLI/UI for people who enjoy running infrastructure
CostApp + $0 storage on machines you already ownFree open source + backend costs (cloud fees or your hardware)

Kopia is excellent at what it does — if you already trust a storage backend and enjoy maintaining repos, retention policies, and verification schedules, it’s a great choice. Recilic exists for everyone who doesn’t want a backend to trust in the first place. The closest Kopia equivalent to Recilic is “Kopia + object lock + a second off-site repo + scheduled verifies + monitoring” — assembled and maintained by you.

Different coin, same metal: Recilic vs. IPFS

IPFS answers “how does the world fetch content efficiently.” Recilic answers “how does my data come back to me after a disaster.” A distribution network vs. a safe.

RecilicIPFS
PurposeBackup: getting your data back after disasterContent-addressed distribution: letting anyone fetch the same content efficiently
Network A closed mesh of machines you own or trustA global public peer-to-peer network — any node can join
Privacy End-to-end encrypted by default; nodes hold unreadable fragments Unencrypted by default — anyone with the CID can read the content; encryption is your job
Persistence The whole product: redundancy, audits, self-healing, restore drills None built in — unpinned content gets garbage-collected. "On IPFS" means fetchable, not preserved
Redundancy Erasure coding: 1.5× space survives 2 losses, placed across failure domains deliberately Full copies wherever someone pins; popular content is everywhere, cold content may exist only on your node
History & versions Immutable, sequence-numbered, key-bound generations — backup semantics built in Content addressing is immutable, but snapshots/retention/timelines are yours to build on top
Verification Active possession audits + rehearsed restores CIDs prove integrity of what you fetched — nothing proves a node still stores it
Repair Lost shards rebuilt automatically from survivors No self-healing — when the last copy disappears, it’s gone

1 · "Fetchable" is not "preserved." The most common IPFS misconception is treating it as a free cloud drive. Content only lives on nodes that actively pin it — unpinned content gets garbage-collected. Persistence means a pinning service (paying and trusting a third party again) or running your own pin cluster. Recilic is the inverse: persistence is the product, and it actively proves it to you.

2 · Opposite defaults: public vs. private. IPFS is built to make content as easy as possible for anyone to fetch — content addressing, dedup, nearest-peer retrieval. Backup needs exactly the reverse: only you can fetch it, and everyone else — including the nodes storing it — sees nothing.

Using IPFS as backup means fighting its defaults — adding encryption, pin management, verification, and repair yourself; finish all that and you’ve roughly rebuilt Recilic. Its ecosystem’s own answer to persistence is Filecoin, whose proof-of-storage incentives are philosophically close to Recilic’s possession audits — except Filecoin pays strangers with tokens to hold your data, and Recilic uses the Macs you and your friends already own. No fees, no tokens, no strangers.

FAQ

Is Recilic a cloud service?

No. There are no Recilic servers and no storage bill. Your files move directly between your own devices over encrypted connections. The mesh is made of machines you and people you trust already own — a cloud VM or S3 bucket can join as just another node if you want an extra off-site copy.

Can people hosting my shards see my files?

No. Everything is encrypted on your Mac before it is sharded and distributed. Nodes store unreadable ciphertext fragments — file names, folder structure, and contents are all encrypted, and no single node (or even several together, below the recovery threshold) can reconstruct anything. Keys never leave your control.

What happens when a node dies or disappears?

The mesh self-heals. Recilic notices a node has gone silent or failed an audit, and rebuilds the missing shards onto the remaining or newly added nodes. With a 4+2 layout you can lose two nodes at once and still restore everything.

What if my main Mac is the one that dies?

That's the scenario Recilic is built for. The encrypted catalog of your backups is itself replicated across the mesh — not just kept on your Mac. On a new Mac, enter your recovery key, reconnect to any enough-of-your-nodes, and Recilic reassembles your files.

How is "backed up" different from synced?

Sync tools tell you a transfer finished. Recilic tells you a restore would succeed: it periodically challenges every node to cryptographically prove it still holds your shards intact, and runs scheduled restore drills. If recoverability degrades, you hear about it before you need it.

Does it protect against ransomware and accidental deletion?

Yes. Backups are versioned and snapshots are immutable — normal writes can't overwrite history, and deletions are delayed rather than immediate. If today's files are encrypted by ransomware or deleted by mistake, yesterday's version is still there to restore.

Isn't iCloud / Google Drive / Dropbox already my backup?

No — those are sync services, and sync's job is to make every device look the same. Delete a folder by mistake, or let ransomware encrypt your files, and sync faithfully propagates the damage everywhere within seconds. Version history is short (typically 30 days, and iCloud has none for most file types), and iCloud's "Optimize Mac Storage" can quietly remove your only full local copy. A mirror is not insurance. Keep them for sync and sharing — Recilic covers what they can't: immutable, permanent, encrypted history of your files, on machines you control.

Why not just pay $9/month for cloud backup?

Services like Backblaze are genuinely good — if you're happy paying rent forever, per computer, and trusting a company both to keep your data and to decrypt it (restoring with a private key means handing them your passphrase that day). Recilic is for the other case: keys that never leave your hands, permanent version history instead of a 30-day window, restores at LAN speed instead of days of downloading, and proof — not promises — that recovery works. If you already own a spare Mac or two, you already own better infrastructure than the rented kind. Why rent what you already own?

When is it available?

Recilic is available now on the Mac App Store — $29.99 one-time.

Backup that proves it can restore.

Download on the Mac App Store