No co-founder. No funding. No backend team. Just a laptop, a lot of failed builds, and a name I gave myself: Akhouri Systems.
Here's everything I've shipped so far โ and what I actually learned building it.
๐ ATLOCK โ the flagship
A Windows security suite that does what most "lock" apps fake:
System Lockdown โ real OS-level enforcement, blocks Alt+Tab, Win key, kills Task Manager on sight
File Guard โ locks files using NTFS Access Control Lists, the same permission layer Windows itself enforces. Not even admin access bypasses it
Password Vault โ AES + PBKDF2-HMAC-SHA256 encrypted, fully local, zero cloud
Intruder Ops โ auto-captures a photo + video the moment someone enters the wrong password
It's fully offline. Independently reviewed and certified "100% Clean" by Softpedia.
๐ ANOTE
A multi-tab text editor with a built-in translator and a lock screen โ built because I wanted a Notepad replacement that didn't feel like it was designed in 2005.
๐งฎ ACALCU
A fully customizable calculator. My first product โ the one that started Akhouri Systems.
๐ท APIC
A local image processing tool โ no cloud upload required to edit or process images.
๐งญ What building this alone actually taught me
Every dependency is a dependency I maintain, forever, alone. That's the real reason ATLOCK doesn't have a server, a dashboard, or a subscription. Fewer moving parts means fewer things breaking with no team to catch it.
The hardest technical problem wasn't UI, it was NTFS ACLs. Getting Windows' own permission system to actually enforce a file lock โ instead of faking it with a password dialog โ meant working directly with the Win32 security API. Getting the save/restore cycle for permissions exactly right took the most iteration of anything I've built.
Shipping solo means every bug is mine, immediately. No triage meeting, no assigning tickets โ if something breaks, I'm the one who finds out and the one who fixes it. That's been the fastest way to actually learn how software behaves in the real world instead of just how it's supposed to.
๐ Where things stand
415+ downloads, zero ad spend
Listed on Softpedia, SourceForge, AlternativeTo, IndiAlt
All builds free, open source, single .exe โ no installer, no subscription
Why I'm sharing this
I don't think age should be the reason to wait before building something real. If you're young and think you're "too early" to ship โ you're probably not. Start now, ship something small, and keep going.
๐ GitHub: https://github.com/Akhouri-Anmol-Kumar
๐ ATLOCK specifically: https://github.com/Akhouri-Anmol-Kumar/ATLOCK
๐ Direct Download Link (ATLOCK) : https://github.com/Akhouri-Anmol-Kumar/ATLOCK/releases/download/v4.0/ATLOCK.zip
Would love feedback from this community โ technical or otherwise. What should I build or fix next?
"We Build What Other Forgot To Fix"
Top comments (4)
Impressive amount of work, especially as a solo developer. One thing Iโd be careful with is the statement that NTFS ACL protection canโt be bypassed even by an administrator. ACLs provide strong access control, but administrators with the appropriate privileges can still take ownership or modify permissions. Clarifying the threat model would make that claim both more precise and more credible.
Reference:
mustafa bro, Thanks
for the detailed feedback, really appreciate it! .You're right, and that's a fair catch โ the wording was too absolute. What I meant was that standard users (non-admin) can't bypass the ACLs without elevated privileges, not that it's unbreakable against an admin with SeTakeOwnershipPrivilege/SeRestorePrivilege. An admin can always take ownership or reset DACLs, so the protection here is really against accidental/unauthorized access by standard accounts, not a security boundary against a fully privileged admin.
I'll update the post to clarify the threat model โ thanks for pointing this out, it makes the claim more accurate. by the way you liked ATLOCK v4?
give some improvements if needed in ATLOCK v4.
Thanks bro โ and yes, v4 has a strong concept, especially keeping everything local and removing external alert dependencies.
My main improvement would be to publish the source code. The repository currently appears to contain only the README, so itโs difficult to review the ACL restore process, vault implementation, emergency bypass, or failure scenarios.
A few other suggestions:
The idea is genuinely impressive. The next big step isnโt adding more featuresโitโs making the security claims independently verifiable.
For Sure.
I will note it down.