Skip to content
AA

Head of AI at NetzeroAl Khobar, Saudi Arabia

Abdulkareem Alramlawi

I build AI that measures the natural world, and products people use every day.

At Netzero I lead the team that measures trees one by one, from a phone photo on the ground and from satellites above, so climate impact can be checked rather than assumed.

Outside work I build with Sada, a small studio making apps for Quran study, reading, family life and desert rescue.

Portrait of Abdulkareem Alramlawi
years building software and AI
5+years building software and AI
engineers in the AI team I built
10engineers in the AI team I built
people using Sada products
300K+people using Sada products
daily rescue requests coordinated by Enjad
6,000+daily rescue requests coordinated by Enjad

Selected work

Experience

  1. Head of AI · Netzero

    Set the AI strategy and built the department from scratch to ten AI engineers, data scientists and pipeline engineers. Own the platform, the roadmap, hiring and the link between field experts and engineering.

    2025 to now

  2. Founding Software Engineer, AI and ML · Netzero

    First AI engineer. Built the infrastructure, data pipelines and the first computer vision models for tree analysis and carbon calculation.

    2023 to 2025

  3. Product Manager · Kar

    Product strategy, market research and roadmap. Ran the engineering team in sprints from requirement to release.

    2023

  4. Product Manager · Enjad

    Led frontend, backend and design teams to deliver the rescue platform, with steady reporting to stakeholders.

    2023

  5. Software Engineer · Garage

    Built a car services product from scratch: backend architecture, database design and documentation.

    2022 to 2023

  6. Undergraduate Researcher · KFUPM

    Cybercrime detection with machine learning: spam and intrusion detection with feature selection.

    2022

Research

Tree diameter from a single photo

Diameter at breast height is the core number in forest carbon. Measuring it from an image needs the trunk width in pixels and the distance to the trunk. The width is solved well by segmentation; the distance from one photo is the hard part. My thesis fine tunes a strong metric depth model on trunk scenes, supervised by a low cost depth camera, so that afterwards an ordinary photo is enough.

Photo
Trunk mask
Depth
Diameter
frames, each label reviewed by hand
4,518frames, each label reviewed by hand
field days, split by day
13field days, split by day
trees measured with calipers
30trees measured with calipers

Master's thesis, KFUPM, 2026. Two papers in preparation.

How I work

Measure first

A claim needs a number behind it, and the number needs a way to be checked.

Honest ranges

Nature varies. A good range is worth more than a precise looking guess.

Rules that travel

Solutions should hold on data they have never seen, not only on the test set.

Small teams, real users

Ship early, sit with the people using it, then improve what matters to them.

Education

  • MSc, Computer Science, AI and Machine Learning

    King Fahd University of Petroleum & Minerals

    Computer vision and deep learning.

    2026

  • BSc, Software Engineering

    King Fahd University of Petroleum & Minerals

    Concentration in AI and Machine Learning. GPA 3.96 of 4.

    2023

Contact