Lagun

A personal AI assistant that learns how you work.

Shape their name, tone, and working style. They'll be in your IM — always available, always remembering your preferences and context across every session.

Identity
Give her a name
Tone

Takes 2 minutes · No setup friction

Traits — optional
Communication
Energy
Interests
Velia
Velia
Playful & SpontaneousDirectAsks questionsSpontaneous

Guess what happened today.

Okay I'm already excited — tell me everything 👀

You're going to love this.

Memory saved · 12 items

Choose your assistant

Start with a preset. Adjust the working style.

Pick one of the starting points below — name, tone, and approach are pre-filled but fully adjustable before you create.

What is a personal AI assistant

How a personal AI assistant differs from ChatGPT and general AI tools.

ChatGPT, Claude, and Gemini are excellent for isolated tasks. But they start fresh every session — they do not know your working style, do not adapt to your preferences, and do not live where you actually communicate. Lagun builds a personal AI assistant that gets more useful the longer you use it.

Learns your working style — not just your last request

Preferences, recurring context, how you like answers structured, what you care about — all stored in visible memory that carries forward. Your assistant starts from where you are, not from zero. Every session.

A working style you actually want

Configure name, tone, and approach before the first task — direct and concise, warm and thorough, or anywhere between. The assistant you set up is the one you get, consistently. Not a generic interface that resets with every visit.

In IM — available without a context switch

After one-time setup on web, your assistant is in Telegram or your preferred IM app. No separate product to open, no login screen to navigate. Reach them the same way you'd message a colleague — in the thread you already use.

Personal Memory

A personal AI assistant that remembers your context — no re-explaining every session.

Every session with a general-purpose AI tool starts from scratch. Lagun stores your working preferences, recurring context, and things you care about in visible memory rows — so your assistant builds familiarity over time instead of treating every conversation as the first.

Context that builds. Memory you control.

  • Preferences, working patterns, and recurring context are stored as visible memory rows — not hidden inside model state.
  • Your assistant references this context in responses without you needing to re-explain it each session.
  • Each memory entry traces back to the conversation it came from — nothing is inferred without a traceable source.
  • You can review, correct, or remove any entry from the web control center at any time.
How the memory center works

IM Access

In your IM — available when you need them, quiet otherwise.

Opening a separate AI interface breaks your workflow every time. Lagun puts your assistant in the messaging app you already use — reachable the same way you'd message a colleague, without a context switch or login screen.

One setup. Always in your thread.

  • Connect a supported IM channel once after your web setup — Telegram or others.
  • Your assistant is then in that thread, with full memory, reachable any time you need a question answered or a task done.
  • No dedicated app to launch, no login screen, no workflow interruption.
  • Return to the web control center to review memory, adjust working style, or manage privacy settings.
Read the IM chat guide

FAQ

Common questions about personal AI assistants.

What is a personal AI assistant?

A personal AI assistant learns your preferences, working style, and recurring context over time. Unlike general-purpose tools that treat every session as fresh, it maintains persistent memory of how you work and becomes more useful the longer you use it.

How is a personal AI assistant different from ChatGPT?

ChatGPT is for one-off tasks with no persistent memory of who you are. Lagun's assistant remembers your working preferences, keeps a consistent personality you configured, and lives in your IM app — no re-explaining context each time.

Can a personal AI assistant remember my preferences?

Yes. In Lagun, every significant preference and recurring context is stored in visible memory rows you can inspect, edit, and manage. Memory persists across sessions and IM channels.

Do I need a separate app to access a personal AI assistant?

No. After web setup, your assistant is reachable in your existing IM app — such as Telegram. The website is a control center for memory and settings; your assistant lives in IM.

What makes Lagun different from other AI assistant apps?

Most AI tools are built for task throughput — they don't know you. Lagun combines configurable personality, transparent long-term working memory, and IM-native access that fits into your workflow without a separate app.

Ready?

Create your assistant. Connect IM. Build something that works with you.

Set their name, tone, and working style. Connect an IM channel. The memory and familiarity build from there — no re-introduction needed, session after session.