ML Engineer (Researcher) is required for a security building solutions company.
Projects:
Development of a web service designed to help sales managers increase sales by providing prompts (to the manager) or directly answering the customer via a bot in messenger on requested products. Working with LMM, it is to investigate and select a model and build a web service interaction pipeline.
The project is aimed at unifying the investigation of various incidents (e.g. shoplifting, product fights, etc.). The project includes recording of the incident, its description, preparation of statistics and recommendations on elimination of possible incidents in the future according to a standard template. The project is on the border of CV and LMM, a series of studies to determine the optimal solution to the problem, to develop a pipeline of work with the model.
Left Behind Detector. Search for abandoned/suspicious items and notify SB of the event. The project is on the border of CV and LMM, a series of studies to determine the optimal solution to the problem, develop a pipeline of work with the model.
Free search in archives of surveillance videos. For example "find a video of two people fighting". It is to make a series of researches to determine the optimal solution to the problem, develop a pipeline of work with the model.
Requirements:
Deep knowledge of matimatics;
Experience in CV development;
Experience in developing NLP systems;
Experience in implementing LLM use cases;
Experience implementing approaches from scientific publications;
Python;
Linux, git flow;
Experience with data collection and partitioning tasks;
Experience developing and maintaining prototypes of business solutions;
Experience in implementing various solutions in CV and NLP (object detection / recognition, tracking, LLM, model optimization);
Understanding of how to optimize the speed of prototypes and inference models.
There is a knowledge of English at the level of reading articles;
Would be a plus:
Knowledge of Machine Learning/Deep Learning;
Experience with opensource LLM models;
Experience in publishing articles at a university.
Format: remote work, flexible schedule.
Time zone: GMT+3.
