Yehuda Levi Group / Sectors / Nº 05 · AI Solutions.
Sector Nº 05 / Applied Intelligence · Agentic Systems

AI that
executes.

We build agentic AI systems that automate, optimize, and operate real-world processes. From data to action — without dashboards, without friction.

Live · AI Solutions.
CORE GRID AGENT forecast · dispatch ASSET AGENT predictive maint. VISION AGENT site safety · CV TRADE AGENT price · arbitrage OPS COPILOT CITY AGENT ▲ AGENTS · 06 LIVE avg latency 42ms · uptime 99.98% ◆ DECISIONS / DAY 2.4M ● HUMAN-IN-LOOP always on AI AGENTS · COMMERCIAL PLATFORM forecasting · maintenance · vision · trading · copilot · city
Sector Nº 05 / Applied Intelligence
PracticeIndustrial AI
Models in prod42
Endpoints2.1M
Domains6
Nº 01 / Approach

How we think
about agentic AI.

We don’t build generic AI. We build the operational layer that makes AI useful.

We design and deploy specialized AI agents for complex operational environments. Our systems are built for Energy, Infrastructure, and Industrial operations. Each agent is designed to make decisions, execute actions, coordinate across systems, and continuously improve through feedback loops.

Our focus is simple: maximize processes through agentic execution. We replace passive analytics with active systems that act.

Nº 02 / Capabilities

Four practices,
one operating discipline.

Nº 01

Agent Design

Domain-specific agents built for execution, not analysis. Focused on decision-making and real-world actions.

  • Task-oriented
  • Multi-agent
  • Planning & tools
Nº 02

MCP Integration

We interconnect platforms through modular context protocols, enabling cross-system orchestration.

  • Orchestration
  • Legacy + modern
  • Real-time sync
Nº 03

Deployment

Agents deployed where operations happen. From on-premise secure environments to edge devices.

  • On-premise
  • Hybrid cloud
  • Edge devices
Nº 04

Control & Observability

Full control over autonomous systems. We do not believe in autonomy without accountability.

  • Human-in-loop
  • Rollback
  • Audit trail
OperationalEfficiencyMaximized
CostReductionAutomated execution
AssetPerformanceContinuously improved
DecisionSpeedZero latency
Nº 03 / System

System
architecture.

Input. Decision.
Action. Feedback.

The Agentic Loop: Input → Context → Decision → Action → Feedback → Continuous optimization. We interconnect platforms through modular protocols.

  • StackEdge + Cloud + Lakehouse
  • RegistryAgent and model DB
  • LayerMCP orchestration
  • ComputeDistributed edge
Step 01Input
Step 02Context
Step 03Decision
Step 04Action
Step 05Feedback
Step 06Optimize
Nº 04 / Use Cases

Selected use cases
delivered or operating.

Case · 01
Case Nº 01

Autonomous BESS
Optimization

◆ AGEERA Platform

An agentic AI system for real-time control of BESS. Continuously monitors energy flows, predicts pricing, and autonomously executes charge/discharge decisions to maximize revenue and extend battery lifecycle.

Costs-30%
ControlAutonomous
Case · 02
ENERGY AGENTS
Case Nº 02

Autonomous PV
Monitoring

◆ Agentic Platform

Agentic AI system monitoring 57 solar projects and 120+ inverters in real time. API-based integration, LLM diagnostics, and AI fault detection to transition from manual to autonomous predictive operations.

Projects57
Inverters120+
“AI that executes. Systems that operate. Results that compound.
Agentic Systems · Operating Principle
Nº 05 / Related

Adjacent practices.

Begin a
conversation about
ai solutions.

HeadquartersDerech Menachem Begin 121
Tel Aviv, Israel
Practice leadAI Solutions Desk
Enquiriesinfo@ylevigroup.com
Request a meeting →