FINANCE-TRAINED. INDUSTRY-FOCUSED. AI-ENABLED.
I’m Seif Hemdan. I turn complex questions into clear specifications, financial models and working tools.
DHAHRAN, SAUDI ARABIA ARABIC / ENGLISH
ONE PERSON. CONNECTED DISCIPLINES.
Finance at KFUPM. Financial analysis and business development. Now, turning questions across materials, energy and AI into useful work.
The person behind the work
ILLUSTRATIVE MATERIALS STUDY
01 / SELECTED WORK
Product specifications, purchasing tools, AI verification and learning design. Open a case to see the work behind it.
A factory-ready brief covering dimensions, energy, heat, protection and certification requirements.
A bilingual tool that compares buyer requirements with offers, flags gaps and drafts clarification questions.
Agreement
isn’t proof.
An internally audited AI prototype, rebuilt around deterministic verification and reproducible replay.
A little more
confidence.
A course and speaking app designed around the everyday conversations my father wanted to have.
SAUDI INDUSTRIAL VENTURE / 2026
Developing a permanent-magnet venture around qualification, traceability and supply assurance. I moved the first milestone from a large factory to one Saudi buyer’s actual specification.
PROJECT RECORD / SEPTEMBER 2026 — Stages and counts reflect my own project record. Detailed source files and correspondence are available on request.
02 / THE CONNECTING THREAD
Follow the questions that connected the work.
Select a project, or play the journey.
The lime path follows the origin story. Dashed links group related explorations; they do not imply a development sequence.
03 / THE TURNING POINTS
I keep the failed results next to the wins. Each constraint changed the next decision.
OOPS / JULY
Launching the original venture required more capital than I had.
Designed one product that could help fund the rest: the ADS charger.
ADS / AUGUST
No contacted factory could confirm the 4.6 mm design without full engineering.
Investigated battery chemistry, heat and materials. The unanswered questions opened a path into magnets and energy.
SAUDI MAGNET PLANT / AUGUST
A 2,000-tonne plant concept needed proven customer demand first.
Rebuilt the plan from the customer backward. That reframing became MAGNEVIA.
MAGNEVIA / SEPTEMBER
No buyer had committed yet.
Reduced the first milestone to one customer’s specification and a sourcing and qualification plan to meet it.
ENTROPY ECONOMY / AUGUST
Factories and data centres needed different products.
Separated industrial energy recovery from data-centre cooling assessment. The latter became the Saudi Thermal Assurance concept.
OCCIR / SEPTEMBER
My internal audit showed that the initial design did not beat a simple baseline.
Rebuilt the system around deterministic verification and replay, then tested the revised prototype.
ENERGY MODELS / SEPTEMBER
A fairer comparison reversed the model’s apparent advantage.
Kept the result and made comparison against a strong baseline a rule for future work.
RESILIENCE RESEARCH / AUGUST
The wearable signals explored did not predict the intended outcome.
Designed a controlled study proposal and sent it to IAU researchers. A proposal, with clinical validation still ahead.
04 / WHAT I BRING
A finance foundation, experience writing supplier specifications, and AI-assisted execution across research, models and software.
Financial models, feasibility studies and decision briefs that show which assumptions matter most.
Plant feasibility → customer-led qualification and supply assurance.
Technical specifications, RFQs and offer comparisons that expose missing terms before the next decision.
Product targets → factory questions, requirements and clarification drafts.
Internal tools, analysis workflows and prototypes, with human review and explicit tests for the important claims.
Research and critique → software, verification and replay.
WHERE I ADD VALUE
Storage specifications. Permanent-magnet supply chains. Cooling assessments. Project economics. These are the questions I’m already working on, with a focus on Saudi industrial localization.
A PROPOSED FIRST 90 DAYS
Map the problem, people, suppliers and data. Deliver a clear brief showing where the value is and which questions still need answers.
OUTPUT / AN OPPORTUNITY BRIEF05 / HOW I WORK
From a phone charger to a magnet venture.
Four moves, repeated.
A dead phone leads to a thin battery. A thin battery leads to materials and the supply chain behind them.
Specs, models and briefs exact enough for a supplier, engineer or buyer to give a useful answer.
Factory feedback, expert outreach and baseline comparisons. A real answer improves the next decision.
Change the plan when the evidence changes. Keep the record so the next project starts sharper.
06 / APPLIED AI
No computer-science degree. My experience with AI grew through the things I needed to make.
“When models agree, that’s a starting point, not proof.”
My first contact was my father’s ChatGPT account: questions, writing and learning what the tools could do.
Research, critique, drafting and building. Compare responses, then check them against sources and real constraints.
AI-assisted code and tests, financial models, interfaces, film experiments and model comparisons on a C++ workload.
07 / THE FOUNDATION
Professional sport taught me to work independently and stay with difficult problems. Building gave that discipline a new direction.
Certificates in progress through Wharton, the University of Illinois and the University of Copenhagen, covering AI for business, strategy, leadership and innovation.
MAGNEVIA, ADS and related research and software.
KFUPM Business School · Market research, proposals and partnership work.
PSA World Tour · Self-funded and self-coached. Ranking #521; university captain and 2023 university championship gold.
King Fahd University of Petroleum and Minerals.
A.A. Turki Group · Variance analysis, budgeting, forecasting and investment support.
KFUPM · Data work across three funded research projects.
08 / THE NEXT CHAPTER
DHAHRAN, SAUDI ARABIA · ARABIC & ENGLISHHave a hard problem in energy, materials, industry or applied AI? Let’s map it, challenge it and build a useful first version.
Start a conversation