GulfMedExams
2026Medical exam prep
AI · Structures, tags, and de-duplicates 15k+ MCQs; powers smart study paths and revision flows.
1M+ MCQs solved
- 20k monthly active users
- 100+ countries
- 15k+ questions in bank
- 40+ documented success stories

Lost Wells
We turn a whole county of old maps into a short, checked, ranked list of wells probably missing from today's records, so crews know where to go first.
Revenue projected full-year, not earned — last 5 months' average, annualized.
Process
You get a checked shortlist of well candidates probably missing from current records — ranked so your crew knows where to go first.
AI does the volume, analysts do the judgment, your crew gives the final answer.
Builds on published research (CATALOG, Berkeley Lab / Ciulla et al., 2024). No partnership claim.
Our team is currently working on this.
Outcome
Lost Wells is in active development. Our team is engaged on old-map ingestion, state-record matching, and analyst review — not on published county results yet.
Our team is actively engaged in this.
Exact locations, ordered so high-confidence sites come first.
Each score ties to map editions, photos, or imagery — not a black box.
The goal is fewer blind visits and less analyst time on the map stack — to be quantified in pilot counties.



Each ranked candidate is meant to ship with map, photo, and imagery evidence your crew can verify before rolling a truck.
Capability
AI is used internally; quality is human-checked.
Track record
$30k+ projected full-year revenueNot earned. Extrapolated from the last 5 months' average.
Medical exam prep
AI · Structures, tags, and de-duplicates 15k+ MCQs; powers smart study paths and revision flows.
1M+ MCQs solved
AI tool discovery
AI · Natural-language search, comparisons, and behaviour-informed recommendations across 1,000+ indexed tools.
1M+ search impressions
AI grading
AI · Scores open responses and surfaces AI-drafted feedback for instructor review.
15k+ assessments
AI presentations
AI · Generates slide structure and draft content from a brief, with human review before export.
1,000+ presentations
Shared scale; counts use the user figures stated on each product card.
GulfMedExams
1M+ MCQs solved
20k users · 100+ countries · 40+ success stories
Leaveit2AI
1M+ search impressions
10k+ users · 1,000+ tools indexed
GradexAI
15k+ assessments
5k+ users · 10+ institutions
AI PPT
1,000+ presentations
AI-drafted decks from briefs
Each block uses that product's headline metric — not compared on one axis.
Launch years match product cards.
People

CEO
AI and machine learning for assessment platforms and applied imaging research.
MIT CSAIL · AI/ML cert. (USA)
PhD · 15+ publications

COO
Microwave sensing, passive RF design, and research-led product operations.
PhD · Leeds (UK)
Postdoc · Gdańsk (Poland)

AI Product Engineer
Full-stack product work across exam prep content, growth, and platform operations.
B.S. · Computer Science
AI product engineering

AI Product Engineer
End-to-end AI product engineering — architecture through launch for lean teams.
B.E. · Electronics Engineering
Ignite AI Wrapper 2025 winner
Legal
Reach us