Deep Learning Engineer – I

Job Description

As a Deep Learning Engineer-1, you will be responsible for developing Computer Vision (CV) – Deep Learning models with a long-term view of integrating them with various Detect products. You will also be expected to take ownership of the data preparation and processing pipelines and innovate on faster ways to use the same. We are looking for Candidates capable of handling oneself in a technically diverse and demanding atmosphere.

Job Responsibilities

1. Train Deep learning models

   ▪ Ensuring that best practices are followed in solving AI problems including thorough problem analysis, prior research understanding, persistent experimentation and documentation, rapid prototyping, and extensive testing.

   ▪ Creating and executing on processes for thorough performance evaluation of the solution against real-world test cases.

   ▪ Benchmarking against competing solutions, prior research and customer’s acceptance criteria wherever applicable

2. Data Management

   ▪ Planning the nature, volume and source of data required to effectively solve the problem.

   ▪ Ideating with relevant people on best ways to source and annotate this data.

   ▪ Following up with Labelling Teams on a continual basis to ensure minimal rework

3. Others

   ▪ Good articulation of your thoughts and the ability to write technical documentation and blog posts

   ▪ To be able to contribute back to open-source libraries and integrate with their testing mechanisms when required

Requirements

• Adapt at modelling Computer Vision problems using Deep Learning frameworks

• Adapt at using git and using IDEs

• Projects, certification, or open-source work showing experience in using deep learning for Computer Vision and Data Management

• 1+ years work experience in computer vision; key areas of interest include object detection, image/ video, tracking and recognition, OCR detection, activity recognition.

Preferred Skills & Knowledge

•  Experience in Docker

• Experience in Nvidia frameworks like TAO, Deepstream, TensorRT

• Experience in fields relating to – Image Stitching, 3-D reconstruction, video analysis, clustering, object detection, image classifications, semantic image segmentation, Large Vision Models (LVM), Large Language Models (LLM), LlamaIndex, LangChain, RAG,

• Knowledge/Experience on deploying programs into compute limited hardware like RPi/Nvidia Jetson

• Strong understanding of theory behind DL techniques such as CNNs, transformers, LLM/LMM supervised and unsupervised Learning, optimization techniques, etc