Our products

The problem. Lab results usually arrive as a page of abbreviations and reference ranges. Mediora explains what each marker measures, what an out-of-range value can mean and when a result is worth discussing with a doctor, in the reader's own language.

What we built. We built the whole product: PDF and photo upload, AI extraction of markers, explanations in three languages, trends across uploads, family profiles with encrypted data, an open marker catalogue and an MCP server for AI agents.

Running it. Mediora is our own product, so SLAtech runs it end to end: servers, releases, monitoring, data security and encryption, the content pipeline and the catalogue, and every new feature.

The problem. Visitors come to a website with simple questions: how much, when are you open, how do I book. When nobody answers right away, they close the tab and the enquiry is lost.

What we built. We built an AI assistant that learns your website and answers from it in Hebrew, Russian and English, linking the source page. In Hebrew the whole widget runs right to left, visitors can leave their details for a person, and the admin panel works in all three languages.

Running it. After launch we run and maintain the platform, refresh the knowledge base when the site changes and track answer quality. We publish the quality benchmark openly, including what does not work yet.

The problem. A client paying for development wants to know what was done, where the hours went and how the budget was spent. The agency loses hours to status reports and meetings. The data exists, scattered across task, time, code and monitoring tools.

What we built. We built a portal that pulls data from the tools the team already uses and answers questions in plain words, with a link to the source. It also covers spec approval via personal links, a signed PDF copy, proposals and issuing invoices in the accounting system.

Running it. The client portal is already in use by our clients. SLAtech develops, hosts and maintains SLAtech Vantage: regular updates, error monitoring and new integrations as our clients need them.

The problem. Running several products in three languages across several networks by hand is slow, and haste breeds mistakes: a post promises a feature that does not exist yet, or goes out twice. We needed posts built only from checked facts, with nothing published without a person.

What we built. We built an engine that writes posts from each product's knowledge base, checks them in three layers, renders cards and clips in the brand's style, Hebrew right to left included, and schedules them. It publishes to Telegram, Instagram, Facebook, LinkedIn and TikTok.

Running it. With Reach we run social media for our own products and for clients, in three languages. It runs on SLAtech's own server. We watch for errors, keep the network connections alive and hold the budget: every paid AI call reserves its cost first, and a single Stop button halts publishing and AI calls.

The problem. New projects appear on xPlace every day, and the freelancer who sends a precise proposal fast gets more replies. Reading every project, estimating scope, pricing it and writing a personal proposal takes hours, and good projects slip by.

What we built. We built an agent that reads the project feed and the site's notifications, filters out projects that do not fit, estimates the scope with AI, computes the price by fixed rules, writes a proposal in Hebrew and submits it. Every submission is reported in Telegram, and one command stops the agent.

Running it. The agent runs on SLAtech's own server. It signs in by itself and stops after two failed sign-in attempts, logs every action and sends an alert when something fails.

The problem. Greetings in a work chat get lost: dozens of similar messages, and two days later nobody can find them. The goal was to gather everyone's words in one place and deliver them on the day itself.

What we built. We built a service where an organiser creates a card and shares one link. Contributors add text, photos, a voice note or video without seeing each other's. At the chosen hour the card opens, with a short film and a song written from the greetings. Yahad is the Hebrew edition, right to left, and its song addresses each writer in the right grammatical gender.

Running it. SLAtech develops and runs the product: hosting, deployments, error monitoring, site content and ongoing improvement.

The problem. Listings tell you where to go, not who with. Events are scattered across many sites and languages, and few people go to a gig or a party alone.

What we built. We built xFind: one listing fed by several sources, a map and a feed filtered by city and date, meetups with a private chat with the host, a bars and clubs directory and a Telegram bot. Marking "I'm going" forms a group automatically.

Running it. After launch we keep running it: event import and translation into three languages run automatically, the site is monitored around the clock and recovers from failures on its own, messages are moderated automatically, and updates ship regularly.

The problem. Women facing difficulties in relationships, at work or with themselves often have nowhere to think things through calmly. Daniela offers warm support in Russian and Hebrew, never diagnoses or decides for her, and points to friends or professionals when it matters.

What we built. We built the mobile app, the server and the admin panel. Every message is risk-checked first, and if there is danger the app shows help contacts for her country. It has a PIN, a decoy PIN, quick exit, conversation rehearsal and memory of her story.

Running it. SLAtech runs the servers, monitors errors and AI spend, keeps help contacts up to date and ships new versions. Sensitive texts are written and reviewed together with a consulting psychologist.