An AI lab building for
everyday life.
We do the research and we ship the things it becomes. The work runs from model behaviour and evaluation through to the interfaces people open without thinking about them.
Research
Model behaviour
We shape how our systems act before we ship them, not after something breaks.
On-device inference
Processing stays close to the user wherever the product allows it.
Evaluation
Every release is measured against safety and quality benchmarks, continuously.
Interaction
Interfaces are built for people first — clear, fast, and honest about what happens to their data.
What the research
turns into.
The same research and judgement, built and operated by us across every product.
Hush
Speak freely. Stay unseen.
Anonymous messaging and Q&A, where the model does the work of keeping a conversation worth having.
- • Anonymous Q&A
- • Model-assisted conversation
- • HushSafe moderation
HushSafe
Safety that understands nuance.
Content classification and review for platforms, with the reasoning behind every decision attached to it.
- • Real-time moderation
- • Appeal flows
- • Enterprise API
HushAI
Intelligence that listens. Judgement you can trust.
A companion that runs on your device where it can, and answers like it has been paying attention.
- • On-device AI
- • Context-aware
- • Safety-aligned
HushDate
Real chemistry. Real connections.
Dating where the matching is a research problem, not a slot machine.
- • Compatibility matching
- • Proximity discovery
- • Safety verification
What we build on
The principles that carry through every product we ship, regardless of how many there end up being.
Research first
Every product starts as a question we can test, not a feature we assume.
Safety by design
HushSafe's review and reasoning runs across every product we ship.
Judgement you can check
HushAI shows its reasoning instead of asking you to trust a black box.
Built for real people
Authentic human connection, through Hush and HushDate.
From the research desk
Notes on what we’re building and why, as it happens.
Latest
All newsTeams building with MatrioshkaLabs
Outcomes from teams building on our products.
hush
Meaningful Connections Through Anonymous Q&A
“I found my community without revealing who I am.”Read the story
10
K+ positive interactions in first month
2,024
Founded
Source: MatrioshkaLabs corporate fact sheet, June 2026
3
Founding team members
Source: MatrioshkaLabs leadership roster, June 2026
What people are saying
“Finally, anonymous messaging that doesn't feel sketchy. The privacy controls are thoughtful.”
“HushAI respects boundaries while still feeling genuinely helpful. On-device mode is a game-changer.”
The Latest from MatrioshkaLabs
Blog posts, news, events, and webinars from the lab.
MatrioshkaLabs at Privacy Summit Bengaluru
Meet the team and see live HushSafe moderation demos.
How We Build Safety Architecture From Research, Not Assumptions
How evaluation-driven moderation keeps users safe without surveillance.
Introducing HushSafe: Content Moderation with Reasoning Attached
Product walkthrough for platform builders evaluating moderation infrastructure.
Trust & Safety in 2026: India's DPDP Act and What It Means for Platforms
Virtual roundtable with platform leaders on compliance, moderation, and user rights.
Frequently Asked Questions
See what we're building
Explore our products or get in touch about partnerships, press, and investor relations.