Engineering Intelligence for Complex Industrial Systems
I design models, optimization and decision systems for real problems in production and operations.
Industrial AI & Optimization Engineer · R&D Leader · PhD Researcher
- System
- Data
- Machine Learning
- Simulation
- Optimization
- AI Agent
- Decision
- Industrial AI
- Optimization
- Agentic Systems
- Digital Twins
- Research
What I actually build.
Eight disciplines, one kind of problem: decisions that must hold up against physics, cost and constraint.
Industrial AI
Machine learning on real production processes: prediction, quality and process behaviour under constraint.
Optimization & Operations Research
Mathematical models — MILP, multi-objective formulations — that turn operational constraints into schedules and plans.
Agentic AI
Agents that read state, use tools and carry a decision loop end to end.
Digital Twins
System representations kept current by online operational data, used to reason about state and future behaviour.
Simulation
Hybrid simulation of industrial systems, paired with learned models for scenario analysis and forecasting.
AI Product Architecture
LLM and multi-agent products designed as whole systems, from data flow and evaluation to the interface.
R&D Leadership
Leading industrial R&D across factories, from framing the problem to a system in production.
AI Strategy & Digital Transformation
Roadmaps, use-case prioritisation, and organisational readiness — including executive-level training for boards.
What the work taught, taught on.
Board briefings, industrial training, university teaching — and an open course anyone can enter.
Fundamentals of AI and AI Engineering
Sessions one and two · rolling intake
Register, answer twenty questions, and see which track fits you.
The course and its survey are in Persian.
Executive AI & Digital Transformation Roadmap
Espandar Civil Industries Co.
Agentic AI Applications in Industrial Operations
Espandar Civil Industries Co.
Industrial Applications of AI in Preventive & Predictive Maintenance
Espandar Civil Industries Co.
Optimization I
Amirkabir University of Technology
Optimization II
Babol Noshirvani University of Technology
Seven steps, from system to decision.
Any industry, any scale, the same loop — and it closes, because each decision changes the system it came from.
- System
I map how your operation actually runs — not how the manual says it does.
- Data
I turn scattered records, sensors and spreadsheets into one stream you can trust.
- Machine Learning
I build models that see quality, failure and cost coming before they arrive.
- Simulation
I test the change on a digital twin first, so being wrong costs nothing.
- Optimization
I find the cheapest plan that still respects every real-world constraint.
- AI Agent
I hand the repetitive calls to AI agents and keep your experts on the hard ones.
- Decision
A decision your team can act on next shift, reasoning attached.
How it works
01 · System
I map how your operation actually runs — not how the manual says it does.
02 · Data
I turn scattered records, sensors and spreadsheets into one stream you can trust.
03 · Machine Learning
I build models that see quality, failure and cost coming before they arrive.
04 · Simulation
I test the change on a digital twin first, so being wrong costs nothing.
05 · Optimization
I find the cheapest plan that still respects every real-world constraint.
06 · AI Agent
I hand the repetitive calls to AI agents and keep your experts on the hard ones.
07 · Decision
A decision your team can act on next shift, reasoning attached.
Physical systems, becoming intelligent.
Learning-Driven Optimization with Agentic AI
Closed-loop production planning: machine-learning surrogates, MILP scheduling and a supervisory AI agent.
- Machine Learning — surrogate models
- MILP — production scheduling
- Agentic AI — supervisory agent
- Operations Research
Products with their own name on the door.
Each one is a complete system doing a whole job — for students, traders, athletes, marketers, and factories.
OstadAI
An AI platform for applying abroad: matching professors and programmes to a CV, outreach, and open positions.
Explore Product
FinAI
Driven by data. Powered by AI.
An algorithmic trading platform and optimisation system with an engine of its own, working with MetaTrader 5.
Explore Product
ARA
AI Beauty Assistant
An AI beauty assistant that analyses skin and guides a personal skincare routine.
Explore Product
GymYar
Strength in motion, together in growth
An AI gym coach, mentor and analyser, paired with access to real trainers.
Explore Case Study
Asr-e Dovom
The Second Age
An AI content factory that runs end-to-end content production for social media and marketing, from research to a finished piece.
Request Private Demo
Espandar Digital Innovation Center
Digital Innovation Center
An enterprise data portal that maps ERP and MES gaps and readies the organisation for digital transformation.
Request Private DemoLocal AI Operator
Running multi-step workflows locally
A local-first operator that drives tools, MCP servers, and applications to carry out complex workflows.
Private SystemRequest Private DemoPredictive Digital Twin
Forecasting how an industrial system will behave
A hybrid simulation and machine-learning twin that represents current state and forecasts future behaviour.
Private SystemDiscuss This Research
Built, and shown on request.
Research that feeds the industrial work.
AI + Optimization + Intelligent Industrial Systems
PhD Researcher
- Institution
- University of Tehran
- Field
- Industrial Engineering
- Branch
- Optimization / AI
- Since
- 2024
Learning-Driven Optimization
Models learned from operational data, built into optimization so it reasons over real process behaviour.
Agentic Decision Systems
Supervisory agents that close the loop between operational state, models, and the people who own the decision.
Predictive Digital Twins
Hybrid simulation and machine learning for representing current state and forecasting future behaviour of industrial systems.
No publications are listed yet; a paper is added only once its status is confirmed.
University of Tehran
PhD researcher, Industrial Engineering
Amirkabir University of Technology
MSc, and teaching
Espandar Civil Industries Co.
R&D Manager, and head of the Digital Innovation Center
Complex systems deserve intelligent solutions.
Engagements usually start with a narrow, well-defined operational problem and grow from there.
This is a fit check, not a free consulting session. If there is a real problem to work on, we agree on the right next step at the end of the call.
Not ready for a call? Ask my digital twin first.Ask the digital twin