Data Scientist (Senior) | Cape Town

We are looking for a Senior Data Scientist to join our team in Cape Town.

About Aerobotics

Our mission is to provide intelligent tools to feed the world. We do this by providing tree and fruit insights to growers in 18 countries around the world, enabled by drone imagery and artificial intelligence. 

We have analysed over 100 million trees and 1 million fruit to date. Our team of 80+ people is based out of Cape Town and we have commercial offices in the USA, Portugal and Australia.

Everything that we do and every decision that is made, is driven by our Core Values.

Context of the role

Provide farmers with accurate yield estimation and management tools that enable them to be as productive as possible.

How we work

The Aerobotics org is split up into customer-focussed streams. Within a stream we have multiple product development squads. Each squad is responsible for owning a specific customer value-proposition from conceptualisation through to delivery, allowing for maximal autonomy and ownership within the team. A squad consists of a Product Manager, Product Designer (if applicable), Engineering Team Lead and technical team members. Our technical team consists of 3 core functions: Software Engineering, Data Science and Drone Engineering.

We use our own combination of agile development methodologies (encompassing our mix of exploratory, discovery & delivery work) to ensure that we are building the best product in an effective manner (without getting lost in process for the sake of it). 

Our tech stack

• We use Python 3 for most of our backend systems. Specific frameworks include Tensorflow for the Data Science work, Django for our APIs and a multitude of internally built libraries containing convenience functions.
• Our frontend web & mobile applications use a mix of Angular, React and React Native.
• Our drones run NVIDIA Jetson Xaviers on-board and are powered by C++, with our own FlightOS operating system.
• We use AWS Cloud using Infrastructure-as-code principles for the bulk of our infrastructure, leveraging serverless computing wherever possible (Fargate, Lambda, Step Functions, Cloudfront, SageMaker etc).
• Continuous integration, delivery and testing is part of every application, ensuring that the technical team can entirely own their work from development through to production. This allows us to push multiple deployments daily, allowing for iterative development & continuous learning. 

Job Purpose

The role is in the Trees Team, which provides farmers with an objective measure of the state of their farm by assessing the health of every individual tree. The farmers use this information to decide on interventions like modifying irrigation lines, applying fertilisers at variable rates, and ordering replacement trees. They also use this information to make business decisions like how much cover to get when insuring their crop. The team provides this information in a way that’s reproducible and scalable, and has set the standard for crop inspections that’s being adopted by both the private sector and government agencies.

As a Senior Data Scientist joining the team, you will be responsible for designing and implementing the models and systems that allow us to provide even richer information at greater scales. For enriching the information that we provide for farmers, you’ll work on questions like “how can we accurately estimate the age and cultivar of a tree from the air?” and “how can we detect water stress?”. For providing this information at scale, you’ll work on questions like “how do we achieve superhuman instance segmentation in a wide range of conditions?” and “how can we infer the presence of small trees in satellite imagery?”. 

More broadly than the team context, you’ll contribute to the technical growth of the company as a whole. You’ll set the standard for technical excellence in data science and will assist others in reaching it through clear technical communication, feedback and mentoring. You’ll work with the Head of Data and CTO in developing and carrying out the company’s long term data strategy.

Required Experience

In order to be effective you will need to have experience with:

• Mechanistic modelling. You’re able to use limited data and the existing body of scientific work to develop mathematical models of the mechanisms that generate given observations. When needed you’re able to approximate solutions to analytically intractable problems through numerical or asymptotic methods.
• Statistical modelling. You’re able to pair appropriate inductive biases with data to make probabilistic claims about new data.
• Machine learning. You’re able to create or improve machine learning models through well-considered choices of the learning objective, architecture, sampling strategy and learning mechanism. You have some familiarity with computer vision tasks like instance segmentation, image registration and depth estimation.
• Software development. You’re able to create robust code that is delivered iteratively.
• Data-intensive systems. You’re able to design, analyse and monitor systems that ingest and process large volumes of data.

You are able to

• Communicate technical concepts clearly, with an excellent understanding of the fundamentals
• Rank solutions based on likelihood to succeed and reliably estimate lead times in projects
• Drive the ideation and implementation of new projects or products 
• Direct and manage the work of others on complex collaborative tasks
• Recognise when teammates need help and foster a culture of learning and self improvement