Skip to content

Internal company communication

A company messenger on your servers

Acquisition plans, salary negotiations, and HR matters circulate today on messengers whose terms of service no one has read. We deploy a dedicated channel: a server on your premises, end-to-end encryption, and an AI assistant on the same network.

In practice

Team-only channel

The communications hub is a server at your premises running your team messaging platform and language model. No account with an external provider and no traffic routed through someone else’s data centers.

This is usually where you handle conversations you don’t want leaving the company: board meetings, HR, projects before they are announced, and correspondence with your law firm. Everything else can stay where it is. We won’t ask you to move the entire company at once.

We select hardware based on how the network operates, not just the number of accounts. A closed node that does not exchange traffic with servers outside the company starts with 6 cores, 16 GB of memory, and 50 GB of system disk space.

This is enough for 500 accounts. For 501–2,500 accounts, the allocation for the conversation server alone increases. When traffic with servers outside the company is allowed, the same node needs 8 cores and 32 GB.

The assistant does not make decisions on its own. It summarizes the discussion and extracts deadlines and tasks. The meeting facilitator decides what goes into the minutes.

Employee correspondence contains personal data, so we configure message retention and log scope to match your security policy.

  • How it works

    Messages, files and threads are encrypted between devices, and the server runs on your premises, not in someone else’s cloud. The language model connected to it summarizes long threads and searches for decisions, but works only within this network.

  • Integrations

    We use accounts from your company directory, so anyone who leaves the company loses access that same day. We connect your calendar, drive and ticketing system as sources for the assistant to read, always within each person’s permissions.

  • Who decides

    Your administrator, not us, sets the keys and how long messages are stored. On a closed network, the database grows by 0.6 GB per active account per year. We access the service only when you grant us access, and every session is recorded in a log you can review.

Problem

Where does the confidentiality of a company chat end?

HEXART solution

Conversations and files stay on a server hosted on your premises, and your administrator holds the keys. The assistant summarizes and organizes threads, while the person leading the meeting approves the decisions.

Discuss a private network

Questions and answers

Frequently asked questions

What hardware does an encrypted business messenger require?

Up to 500 accounts, a closed node is sufficient: 6 cores, 16 GB of memory and a 50 GB system disk. For 501–2,500 accounts, allocate more resources to the call server. A node that exchanges traffic with servers outside the company needs 8 cores and 32 GB.

Do you need to move your entire company to a new messaging platform?

No. It usually covers the conversations you do not want leaving the company: board discussions, HR matters, projects before announcement, and correspondence with legal counsel. Everything else can stay where it is.

Who holds the encryption keys?

Your administrator. They also set how long messages are stored. We access the service only after someone on your team grants us access, and each access is logged.

What does the AI assistant do within this network?

Summarises long threads, finds decisions, and extracts deadlines and tasks. Always within each person’s permissions and strictly inside the network. The meeting host approves what goes into the minutes.

No-obligation call

Schedule a workshop with Paul

Choose thirty minutes or a full workshop. Book a time slot in the calendar, receive immediate confirmation.

Paul Lazniak

HEXART Founder

Contact

Let us talk about your project

We will show you a working product before we start talking about it. Let us know what you need: a film, an XR environment, an AI system, or brand identity.

Write or call

A few sentences are enough: what needs to be created, by when and for whom.

Address
Aleja Zwycięstwa 96/98, 81-451 Gdynia

Deployments catalog

View the full catalogue
  • On-premise server knowledge base (RAG)

    Search contracts, technical documentation, and correspondence via chat when company policy prohibits sending them to the cloud. The model and index run in your rack cabinet. A single machine handles 5 to 20 concurrent simple queries within a 100 to 1,000 ms window. An additional server doubles that capacity.

  • Dashboard and executive assistant

    Arguing over whose report is accurate costs more than reporting itself. The dashboard gathers data from your existing systems, displaying the source and timestamp for every metric. It refreshes up to once per minute, depending on API limits.

  • Local AI server (Blackwell, 128 GB)

    Sending company documents to an external API can be unacceptable legally or simply out of caution. We deploy an on-premises machine with 128 GB GPU memory running open-source language models with no cloud connection.

  • AI agents for chat and phone

    A large share of support tickets are repeated questions about order status, invoices, and delivery times. An AI agent answers them directly from your documentation, resolving 30–60% of such inquiries in our deployments. It routes the rest to a human agent along with full conversation context.

  • SEO blog content generator

    Revenue-driving expertise stays locked inside your team because nobody has half a day to write an article. We capture it once, and the system drafts search-optimized articles from it.