Deligo is an AI-first company: machine learning is not just an afterthought for us, but it’s at the heart of our product. We build the hardware that solves the customer problem, this allows us to develop a unique and proprietary dataset and build models that drive high-value functionality. Currently we are building visual product recognition for cafeterias. One of the challenges we face is that previously unseen items routinely appear during operation, thus we need to go beyond conventional ML techniques.
Today we serve dozens of cafeterias across Europe with an ambition to scale globally and build solutions for other retail use cases. We are a repeat founding team with VC backing.
Sounds interesting? Get in touch with us!
What to Expect
The Role:
You will be part of a team of developers and machine learning experts working to design, implement, maintain, and operate systems for dataset generation, data exploration, and other supporting applications necessary to advance the state of the art in perception for AI Checkout.
What You’ll Do
Responsibilities:
- Design and implement large-scale data processing pipelines that handle gigabytes of AI checkout data and human labels.
- Support the training of neural networks by performing tasks such as optimizing code & data formats for faster dataloading, orchestrating data auto-labeling jobs, etc.
- Improve our trigger infrastructure, that automatically detects corner cases and feeds it into the Data Engine.
- Building tools to visualize and summarize the statistics of training data to support efforts to root-cause poor label quality and improve the training procedure of our neural networks.
- Support the deployment of trained models into production and help scale our Neural Inference microservice.
What You’ll Bring
Requirements:
- Strong experience with Python and software engineering best practices.
- Experience in system-level software and resource utilization, parallelism and Python multiprocessing, and being able to debug problems across the stack.
- Experience working with the underlying storage formats for various kinds of data (e.g. images, videos, depth arrays, human labels, etc) and databases (e.g. SQL, Redis)
- Familiarity with python libraries such as numpy, matplotlib, pandas, flask, jupyter.
- Knowledge of cloud computing platforms such as AWS, GCP or Azure
Considered as a plus:
- Understanding of machine learning concepts and experience working with deep learning training frameworks, ideally PyTorch
- Prior experience with object detection and instance segmentation datasets (e.g. MS COCO)
- Knowledge of MLOps practices and frameworks (e.g. DVS, Weights&Biases)
- Experience in deploying and scaling microservices (Docker, Kubernetes)
What We Offer
- An opportunity to work in a small, multidisciplinary team to build the future of checkout.
- Joining at an early stage: you will have an outsized impact on the company
- Competitive salary and options in a small, rapidly scaling company
Additional Information: A few practical details about the role
- Start: As soon as you are ready!
- Location: Budapest
- Contract: Permanent contract
- Compensation: Competitive package based on experience