AI-generated renderP-01 · Streamlit ML + WITSML
BitWearPredictor
A machine-learning app that predicts drill-bit wear from WITSML data.
- Streamlit
- Machine Learning
- WITSML
Tehran · Petroleum Engineering · Applied AI
AmirMahdiTajik
I build machine-learning and retrieval (RAG) tools for drilling data: bit-wear prediction from real-time WITSML, and LLM systems grounded in engineering documents.
Graduated B.Sc. in Petroleum Engineering · Sharif University of Technology
- 01 · A PDC drill bit, AI-generated render
- 02 · Steel dissolving into data, AI-generated render
- 03 · A neural core, AI-generated render
- ROP
- 0m/h
- MSE
- 0ksi
- Hook load
- 0klbf
- Tokens/sec
- 0tok/s
Simulated · illustrative values
01 / About
Engineer by training. Data scientist by habit.
I work where drilling physics meets machine learning.
- Industry internships
- 2
- Petroleum Engineering, graduated
- B.Sc.
Roles
Jul 2025 — present
AI & Data Analysis Intern
Pasargad Energy
Jun 2025 — present
Vice President, Petroleum Engineering Society
Iran Elites Network
Ongoing
Member of the General Council
Petroleum Engineering Society · Sharif University of Technology
Aug — Sep 2025
Drilling Office Intern
Pasargad Exploration & Production
02 / Core Lab
AI core and lithology synthesizer
Pick an interval of a simulated core. This scripted mock-up shows how a domain-specific RAG pipeline with a local LLM could retrieve analog formations, describe the facies, estimate porosity and permeability trends and draft an engineering summary.
Simulated example. The core, analogs and measurements are illustrative, not real data.
03 / Focus
Where the work lives
Move the pointer over the cards
01 · LLM·RAG
LLMs & Retrieval
Domain-specific language models grounded in real engineering documents.
- RAG architectures: chunking, hybrid retrieval, reranking
- Evaluation of grounded answers against source reports
- Adapting language models to oil and gas work
02 · WITSML
Drilling Engineering
Optimization and analytics on real-time drilling data.
- WITSML ingestion, cleaning and depth/time alignment
- Bit wear modeling: physics + machine learning
- ROP, WOB and torque analytics for drilling optimization
03 · ML·PY
Applied AI & ML
Practical pipelines from raw data to deployed insight.
- Python, pandas, scikit-learn, gradient boosting
- ETL automation with n8n and PostgreSQL
- Turning models into tools engineers trust
4
Hands-on roles
Data in, decisions out
04 / Playground
Play with the work
Three small playable models. All are illustrative: simulated data and hand-built examples, with nothing real inside.
05 / Projects
Case studies
- AI-generated render
AI-generated renderP-02 · Retrieval-augmented generation
Gas Lift & Artificial Lift RAG Platform
A RAG platform for gas lift and artificial lift knowledge.
- RAG
- Gas lift
- Artificial lift
AI-generated renderP-03 · On-premise AI · n8n · Ollama
Privacy-First Enterprise Automation & Local Document Intelligence
Enterprise system automation with n8n and confidential document extraction, running entirely on local infrastructure so no data leaves the company network.
100%on-premise
- n8n
- Ollama
- ChromaDB
- On-premise
- Local LLM
06 / Interactive
How retrieval works, live
Type or pick a question. Every chunk is scored, and the top matches are highlighted as the context a language model would receive.
07 / Education
Foundations
2022 — 2026Graduated
B.Sc. Petroleum Engineering
Sharif University of Technology
Thesis
Machine Learning-Based Bit Wear Prediction in Drilling Operations
Supervisor: Dr. Saeid Jamshidi
Certifications
Machine Learning with Python
Maktabkhooneh · 2025
Python Programming (Advanced)
Maktabkhooneh · 2025
Python Comprehensive Course
Maktabkhooneh · 2025
08 / Work with me
Let's build something that reads the ground.
Open to research collaborations, internships and applied AI projects in energy.
Connect on LinkedInWorked and studied with
- Sharif University of Technology
- Pasargad Energy
- Pasargad Exploration & Production
- Petroleum Engineering Society