Technological tasks
The Tech-tasks module is tasks related to the development, implementation and improvement of new technologies, products and processes. If you have development, implementation and other needs you can post information below.
It is necessary to develop a custom SDK for working with the map based on the MapLibre Flutter SDK
Description: You need to fork the open-source SDK MapLibre for Flutter and make changes to implement the functionality of natively moving the placemark fixed in the center of the map. When moving the map, the placemark should remain fixed in the center of the screen, and not "snap" to the coordinates on the surface. The coordinates should be updated as the map moves, and the map should not hang or lose performance. Functionality requirements: Fork the existing MapLibre SDK for Flutter. Add the possibility of a fixed central marker (anchor marker in center). When dragging the map, the placemark remains visually stationary, only the coordinates are updated. The card's performance should not deteriorate: no lags, freezes, or FPS drawdowns. The implementation should be as native as possible (not through UI crutches on top of the map, but through SDK refinement).
Decision acceptance deadline
26.01.26 (inclusive)
Preferred systems
Internet of ThingsNumber of applications
2
Integration system
Successful integration of the new calendar platform with CRM systems, partner services and clinic servers
Decision acceptance deadline
26.01.26 (inclusive)
Preferred systems
Processing and storage of Big DataNumber of applications
2
AI-adaptive test generator for teacher training for NCT/OSP
The system must be able to: Generate test questions automatically based on the available knowledge base (NCT/OSP formats, lessons, methodological materials). Create full-fledged test blocks, including: possible answers, the correct answer and a brief justification, the level of difficulty, and thematic binding. Work in adaptive testing mode — increase/decrease the difficulty depending on the results. Create personal test kits based on the strengths/weaknesses of a particular teacher. Support Russian and Kazakh languages. Integrate with Mugalim Online LMS and statistics system. Functional Requirements (MVP): Loading and structuring the database of tests/topics An AI model for generating questions on a topic Classification of complexity and competencies Explanation generation (“why this is the right answer”) Adaptive CAT algorithm (Computer-Adaptive-Testing) Personal Progress Tracker + recommendations
Decision acceptance deadline
26.01.26 (inclusive)
Preferred systems
Other technological solutionsNumber of applications
4
An AI module for the platform that automatically analyzes test results
The goal of the project is to develop an intelligent module for the Mugalim Online platform, which automates the preparation of teachers for NCT/OSP and helps them build a personal professional growth trajectory. As part of the project, it is planned to create an AI system that will: Analyze the results of teachers' tests (NCT/OSP and internal Mugalim Online tests), taking into account topics, competencies and the level of complexity of the questions. Create a personalized professional development roadmap based on the strengths and weaknesses of the teacher: recommended courses, lessons, tests, practical tasks, deadlines and priorities. Provide an AI simulator of the qualification test: adaptive test sessions that are as close as possible to the real NCT/OSP; dynamic selection of questions based on the level of knowledge; detailed feedback and error analysis. Support the Russian and Kazakh languages so that teachers can study in a comfortable language environment. Integrate with the existing Mugalim Online platform (teacher's personal account, test database, lessons, statistics). The result will be a single AI module that not only tests knowledge, but also helps teachers see their progress, understand what needs to be improved, and purposefully prepare for certification.
Decision acceptance deadline
26.01.26 (inclusive)
Preferred systems
Other technological solutionsNumber of applications
3
Development of a facial aliveness check module (anti-spoofing)
The goal of the task is to develop an anti-spoofing module for use in mobile applications developed on Flutter, with support for iOS and Android. It is necessary to develop a software module that automatically determines whether a live person is in front of the camera of a mobile device or whether substitution is used (photo, video, screen of another device, etc.). The module should work transparently for the user, without performing additional actions (head turns, facial expressions, gestures, etc.), and be optimized for fast and stable operation on massive budget smartphones and tablets.
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Other technological solutionsNumber of applications
3
Development of binding software for integration with Hikvision time tracking and access control terminals.
The purpose of the task is to develop a binding software for the integration of a time management and access control system with Hikvision terminals. It is necessary to develop a software module (middleware) that ensures the receipt, processing and transmission of data (user data, photos of persons, attendance data) between Hikvision terminals and external information systems located on the Internet (not on a local network). The solution should provide stable data exchange, correct interpretation of events and the ability to scale with an increase in the number of devices.
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Other technological solutionsNumber of applications
4
Development of a plug-in for the iiko system for cash terminals
Creation of a plug-in for iiko cash terminals, which provides direct data exchange between the cashier and external (third-party) information systems with the ability to receive and process responses in near real time. The task is to implement a software module in c# .net that would allow: transfer data on cash transactions directly from the iiko cash terminal to external systems; receive responses from external systems and use them in the logic of cash transactions; circumvent the limitations of the standard iiko server integration in cases where serial data transmission through the iiko server is impossible or incorrect from the point of view of the business process.
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Other technological solutionsNumber of applications
0
ZERDE (Zerde)
Objective: To develop and implement a local intelligent search engine (Enterprise Search) that provides instant access to corporate knowledge in a secure closed loop of the organization. Description: Creation of a platform based on RAG technologies that indexes internal sources (EDMS, files, databases) and generates accurate responses to employee requests. The product will solve the problem of wasting time searching for information by combining data into a "single window" with strict observance of access rights and language specifics (kaz/rus), guaranteeing complete confidentiality on the customer's own servers.
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Neurotechnology and artificial IntelligenceNumber of applications
7
A test case for a Data Engineer job
The company receives data from multiple sources, generates business metrics, and uses AI agents to automate processes. It is necessary to design and implement a basic data and AI pipeline Part 1. There is a REST API at https://api.example.com/ads The API returns objects with the following fields date campaign_id spend clicks It is necessary to implement a Python script that performs the following actions: Gets data for the last 7 days Handles API errors correctly Logs execution statuses Saves data to a CSV file or to a database Part 2. There is an orders table with the following fields: order_id user_id order_date revenue It is necessary to write SQL queries that calculate the following indicators: Revenue by day Number of orders by day ARPU Number of repeat orders Part 3. It is necessary to describe the DAG in Apache Airflow, which consists of the following steps: Getting data from the advertising API Getting order data from a database Transformation and calculation of daily metrics Uploading aggregated data to the storage You need to set up: Launch schedule Retreats Logging and error handling Part 4. It is necessary to describe the structure of the daily_businessmetrics table The table should contain the following fields: date revenue orders_count ad_spend roi It is necessary to describe the formulas for calculating each indicator Part 5. It is necessary to describe the concept of an AI agent that uses data from the daily_business_metrics showcase and performs one of the following business tasks: Control of revenue anomalies Monitoring the effectiveness of advertising Notification of a drop in ROI Generating daily business reports You need to describe: What data does the agent use? How often does it work What decisions does he make Where is sending the result of the work It is necessary to develop criteria for evaluating the agent's performance.
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Intelligent control systemsNumber of applications
3
"Intelligent routing and verification of documents in the logistics chain" Description
The company processes hundreds of shipments per day through several shipping companies and warehouses. Each document (invoice, bill of lading, bill of lading, quality certificate) must be checked for compliance with the following requirements: • the correctness of the details (order number, supplier's INN, product code); • matching the actual delivery route with the planned one; • Control of departure dates and statuses. Goal: • Reduction of manual control by 80% • Exclusion of "lost" or incorrect documents • Optimization of logistics processes based on data from EDI
Decision acceptance deadline
22.01.26 (inclusive)
Preferred systems
Intelligent control systemsNumber of applications
7
