01 / 08
Engineering Studio

AI agents for systems that cannot stop

We engineer software, electronics and autonomous control loops

Sensors128
Agents06
StatusNormal
HaClever
02 / 08

Evidence instead of promises

The team has been through 20 Russian and international hackathons and took 15 prize places — from Junction in Helsinki to winning Digital Breakthrough. Below is a verifiable chronology with sources.

ENGINEERING ARCHIVECHRONOLOGICAL LEDGER
  • 2018

    The team's first hackathon — Saint Petersburg

    Prize for the best technological solution: a portable computer-vision camera assembled over a weekend.

    COMPETITIONSOURCE
  • 2018

    Local Hack Day / Impossible.Hack — organizers

    Dmitry Ichetkin and Emil Sataev organized a local hackathon in Zelenograd: about 60 participants, from prototypes to working demos in one day.

  • 2018

    Junction, Helsinki — 3rd place in the Blockchain track

    Europe's largest hackathon: third place in the track for a decentralized asset registry on Ethereum.

    COMPETITIONSOURCE
  • 2019

    Skoltech / Akado Telecom hackathon — 2nd place

    A system automating replies to users on social networks: NLP request classification and suggested operator responses.

    COMPETITIONSOURCE
  • 2019

    IoT hackathon Skoltech / Sberbank / Huawei

    Special nomination for a solution on the Huawei OceanConnect platform: device telemetry and IoT fleet control scenarios.

    COMPETITIONSOURCE
  • 2019

    Digital Breakthrough finals — winner

    Won the Association of Volunteer Centers nomination: a media platform with a video-generation neural network. Prize — 500,000 ₽. Finals with 3,000+ participants.

    COMPETITIONSOURCE
  • 2020

    48-hour MIET Hackathon — partner

    HaClever acted as a partner and authored two applied challenges for student teams; mentoring at the intersection of hardware and software.

    PARTNERSHIPSOURCE
  • 2023

    Publication: multilevel a-Si memristive structures

    Co-author of a scientific paper on multilevel amorphous-silicon memristive structures — a joint work of MIET and the Kurchatov Institute.

    PUBLICATIONSOURCE
TEAM RECORD
20

HACKATHONS

participation and organization

15

PRIZE PLACES

as a team

03 / 08

One team — the whole circuit

Full cycle: from research and software to hardware and on-site deployment

01

Software

Intelligent agents, digital twins and industrial telemetry

AI-Agents

Autonomous agents for control and optimization

Industrial Backend

Reliable backend for mission-critical systems

Telemetry

Data collection and processing in real-time with edge-aggregation

Digital Twins

Virtual models for simulation and forecasting system behavior

03

Research

Own research and algorithms for complex industrial tasks

Memristor Computing

Energy-efficient computing on new physical principles

Anomaly Detection

Early stage detection of abnormal operating modes

Causal Analysis

Understanding causes of deviations and decision support

04

Deployment

We implement turnkey solutions and ensure their continuous operation

Commissioning

Commissioning and integration with existing infrastructure

Observability

Full transparency: metrics, logs, tracing, alerts

Secure Updates

Secure updates without downtime and with integrity control

02

Hardware

Reliable hardware platform for collection, processing and control at the device level

  • Embedded Controllers

    Industrial IoT gateways and edge controllers

  • Sensors and Meters

    Pressure, flow, temperature and electrical metering

  • Edge Nodes

    Fanless industrial computers for field deployment

  • Robotics & IoT

    Sensor networks, robotics integration and edge computing

All directions
04 / 08

AI agent for electrical equipment

A multi-agent system continuously analyzes equipment status, coordinates actions and ensures safe execution of decisions under operator control.

Input

Sensor network
PLC / RTU

Analysis / Orchestration

Diagnostics Agent
Data Agent
Supervisor Agent

Action Planning

Alarm Agent
Maintenance Agent
Operator Approval
Equipment Action
Immutable Audit LogAll decisions, data, approvals and actions are tamper-evident and traceable.

Early anomaly detection

Detects deviations early and focuses on real risks.

Explainable recommendations

Shows causes, impact and alternative action scenarios.

Controlled execution

Actions are executed safely, with operator approval and full audit.

05 / 08

From signal to verifiable action

HaClever

01 Observe

sensors and telemetry in real time

Anomaly detected

Root cause: insulation degradation

Confidence: 0.92

02 Diagnose

AI identifies anomaly and root cause

Recommendation

Isolate Transformer T-2 due to insulation risk

Risk: High

03 Decide

supervisor agent recommendation

Confirm critical action

Engineer: A. Petrov
Confirm

04 Approve

engineer confirms critical action

Command executed

Equipment: CB-17

Action: Open

2026-07-20 09:41:03 UTC

Audit log · sealed

05 Act & Learn

equipment command and audit log

Critical actions remain under human control

MISSION // RELIABILITY

PRINCIPLE // VERIFY BEFORE YOU ACT

06 / 08

SELECTED PROJECT

AI diagnostics for distribution equipment

CHALLENGE

Early defect detection without production downtime

SYSTEM

AI platform for telemetry and equipment behavior analysis

STACK

Edge controller, ML models, SCADA, MQTT, OPC UA

OUTCOME

Continuous monitoring and reduced unplanned outages

ENGINEERING ARCHIVE
  • 2024

    Industrial Telemetry Platform

    Sensor data collection, aggregation and visualization: edge gateways, stream processing, operator dashboards and integration APIs.

    SOFTWARE
    SOURCE: GitHubActive
  • 2019–2024

    Web platforms: Lectr, Vimbo, Prizma

    Subscription content platforms with an audio player, an audio publishing house with e-commerce, and a Telegram end-to-end analytics SaaS.

    SOFTWARE
    SOURCE: lectr.ruIn production
  • 2024

    Edge Controller Prototype

    An industrial controller for local telemetry processing and running ML models next to the equipment.

    HARDWARE
    SOURCE: InternalPrototype
  • 2023

    Memristive Structures Research

    Multilevel a-Si memristive structures for analog computing — jointly with MIET and the Kurchatov Institute.

    RESEARCH
    SOURCE: Ioffe InstitutePublished
  • 2019

    Digital Breakthrough — finals winner

    A media platform with a video-generation neural network; won the Association of Volunteer Centers nomination, 500,000 ₽ prize.

    EVENTS
    SOURCE: MIETWinner
  • 2020

    MIET Hackathon Partnership

    Authored two applied challenges and mentored student teams at a 48-hour hackathon.

    EVENTS
    SOURCE: MIETCompleted
07 / 08
Pt (TE)80 nma-Si (L3)25 nma-Si (L2)20 nma-Si (L1)30 nmTiN (BE)40 nmSiO₂200 nmSEM (BSE) · Tilt 52°HaClever Research · a-Si stack100 nmI–V · L3Vlog I

Research

We study matter. We build working systems.

Together with MIET and the Kurchatov Institute, we research memristive structures and their integration into AI systems and industrial electronics.

2023

Multilevel a-Si memristive structures for analog computing

DOI ↗

Team / Studio

Dmitry Ichetkin

Dmitry Ichetkin

founder / engineer

Emil Sataev

Emil Sataev

co-founder / AgentOps

Sasha Grigory

Sasha Grigory

head of frontend

research and developmentelectronics and embedded systemsmaterials and processesengineering solutions
08 / 08

Have equipment and a complex task? We'll assemble a verifiable circuit.

Tell us about your task — we'll respond within one business day.

The more details, the more accurate the response.