Data Scientist & AI/ML Engineering Intern

About DeepLeaf

DeepLeaf is a Moroccan agritech company building agentic AI for agriculture. Our models see a crop and reason like an agronomist, helping farmers, cooperatives, and agricultural partners make timely, practical decisions across many crops, conditions, and languages. Learn more on the DeepLeaf Agronomist and Data Hub pages.

About the role

We're hiring a Data Scientist & AI/ML Engineering Intern to help train, evaluate, and ship the models at the heart of DeepLeaf. You'll work across vision, language, and voice — from dataset to training run to rigorous evaluation to deployment — on hardware that ranges from supercomputers to tiny edge devices in the field.

What you'll do

• Train models. Build and fine-tune models — including VLMs (vision-language), LLMs, and computer vision models — for real agricultural tasks.
• Evaluate rigorously. Design datasets, metrics, and evaluation pipelines; measure quality honestly and catch regressions before they ship.
• Work across modalities. Vision, language, and voice — including data processing, augmentation, and multimodal experiments.
• Run at scale. Launch and manage training/experiments across HPC / supercomputers (e.g. CINECA Leonardo) and cloud (Azure, AWS, GCP).
• Deploy to the edge. Optimize and run models on edge devices (NVIDIA Jetson, Raspberry Pi) for in-field inference.
• Keep it reproducible. Track experiments, version data and models, and document what worked and why.

Who we're looking for

We care about curiosity, rigor, and the ability to learn fast. You should have hands-on experience (through projects, research, internships, or serious personal work) with several of the following, and eagerness to grow into the rest:

• Solid Python and a deep-learning framework (PyTorch and/or TensorFlow).
• Training and evaluating models — you understand metrics, baselines, overfitting, and honest evaluation.
• Some experience with VLMs and/or LLMs (fine-tuning, prompting, RAG, or evaluation) and computer vision.
• Exposure to any of: HPC / supercomputers (CINECA Leonardo or similar, SLURM), cloud (Azure / AWS / GCP), or edge (Jetson / Raspberry Pi).
• Bonus: speech/voice models, data pipelines, model optimization/quantization, MLOps, or a research publication.

Eligibility

Aimed at students and recent graduates. You should be currently enrolled in a university programme (bachelor's or graduate), in your final undergraduate year, or a recent graduate (within about a year).

Languages

• English — required (our research, code, and docs are in English).
• French and Arabic/Darija are a plus.

Location & logistics

• Full-time internship, 6–9 months, in person at the DeepLeaf office in Sidi Bennour, Morocco.
• You're expected to relocate to Sidi Bennour for the internship; living costs there are low.
• Paid internship — a monthly stipend, all-inclusive, intended to cover your stay (accommodation) and local travel.

Good to know

• Confidentiality and responsible data: you'll handle company, partner, and farmer information with care, consent, and confidentiality.
• Equal opportunity: we welcome applicants of all backgrounds and strongly encourage women and applicants from under-represented groups to apply.

What you'll gain

Real experience training and shipping modern AI — from supercomputer to edge — on problems that matter, with mentorship and a strong reference, portfolio, and network for your next step.

How to apply

Use the Apply for this role button on this page. Share links to your GitHub, papers, Kaggle, or model/eval projects — we love seeing your work. We review on a rolling basis.