Yaremchuk / Engineering brand
Engineeringintelligent systems.
Software · AI · Devices · Cybersecurity
One engineering brand across software, AI, connected devices and cybersecurity. Explore the technology layers, the practical work and the directions taking shape.
Start with the whole system
A useful system starts with the problem it needs to solve. Architecture connects data, software, intelligence and security around that goal.
- 01Explore
- 02Design
- 03Build
- 04Validate
Define the outcome and constraints, choose the right components, then build and validate the connections.
Technology stack
- System layers
- Interfaces
- Services & APIs
- Data flows
- Intelligence
- Custom models
- Cloud AI
- Edge processing
- Design criteria
- Latency
- Access boundaries
- Maintainability
Turn video into useful signals
Detection and tracking connect a camera feed to useful events. Processing close to the camera helps the system respond in real time.
The focus: reliable detection, low latency and clear behavior when the connection drops.
In development: VisionOps.
Technology stack
- Video input
- RTSP
- ONVIF
- Perception
- OpenCV
- YOLO
- Object tracking
- System connection
- Edge AI
- MQTT
- Event streams
Connect perception to action
Research into systems that sense their surroundings, interpret signals and turn decisions into physical action.
Sensors, perception and control form one feedback loop, with predictable behavior when information is incomplete.
Research direction: robotic scenarios.
Technology stack
- Research · input
- Sensors
- Computer Vision
- Research · logic
- ROS 2
- Tracking
- Edge AI
- Research · action
- Control loops
- Feedback
Make device data understandable
Vehicle signals become clear explanations that help a person understand a fault and decide what to check next.
Combine the fault code with context, explain uncertainty and make the next step clear.
Technology stack
- Device layer
- ELM327
- OBD-II
- CAN Bus
- Interpretation
- Fault codes
- AI
- Vehicle context
- Product layer
- Mobile UX
- Diagnostic reports
- Next steps
Build around real user needs
Product design, e-commerce and behavior analytics connect what people need with how a digital product works.
Map the user journey, find friction and use data to guide improvements.
Technology stack
- Experience
- Product UX
- User journeys
- E-commerce
- Measurement
- Events
- Heatmaps
- Funnels
- In development
- ClickLens
- AI Insights
Match intelligence to the task
Custom models, ready-made AI and cloud platforms offer different paths to a useful result: predictions, recommendations and intelligent product features.
Choose the model and infrastructure for the data, quality target and cost. Evaluate on new data and improve through feedback.
Myron and Auroxent are in development, including ML experiments for trading scenarios on Binance.
Technology stack
- Custom models
- Prediction
- Recommendations
- Model evaluation
- Cloud options
- Google Cloud / Vertex AI
- Model APIs
- Managed training & inference
- Product integration
- Data pipelines
- Quality metrics
- Feedback loops
Make tools work together
AI, APIs and MCP connect tools into workflows for content, localization and advertising operations.
Define each step, control access and make failures and human review part of the process.
In development: RePostIQ and AdHatch, an MCP for advertising networks.
Technology stack
- Connections
- APIs
- MCP
- Tool access
- Workflow logic
- Routing
- Orchestration
- Human review
- Applications
- Content
- Localization
- Advertising networks
Build trust into the connections
Cybersecurity is a growing direction of the brand, focused on protecting communications, data and the connections between system components.
Develop security alongside the architecture: understand what needs protection, define access boundaries and make controls verifiable.
Practical experience: an Asterisk communication channel with SRTP media encryption. The direction will expand through further projects.
Technology stack
- Applied foundation
- Asterisk
- SRTP
- Media encryption
- Development focus
- Secure communications
- Data protection
- Access boundaries
Yaremchuk
Many disciplines. One system.
Engineering & Intelligent Systems
Software, AI, devices and cybersecurity develop under one brand. The capabilities grow with new projects and collaborations, while the system remains the common point of focus.