Deep Learning Engineer
Soothsayer Analytics is seeking Deep Learning Engineer to research, design, implement, optimize and deploy deep learning models for variety of exciting projects including NLP, Computer Vision, Time Series.
Education and Experience:
PhD in computer science, Physics, Statistics, applied Mathematics, or other quantitative/ Computational discipline or Master's with over 2 year of real time relevant experience in deep learning is a MUST.
Overall experience: > 3 years
- Develop state-of-the-art and novel deep neural network architectures using various types of data modalities such as high throughput behavior data, images, and text data
- Develop training & testing pipelines to assess the performance of architectures on relevant data processing tasks
- Keep up with deep learning literature and research publications to implement the latest techniques into our networks and pipelines
- Develop machine-learning algorithms on a breadth of software frameworks (Keras, TensorFlow, Torch) and deploy on a diversity of hardware platforms
- Create or guide data labelling requirements as needed using different tagging software
- Pre-process, clean and wrangle data using Pyspark/Python as required
- Work on distributed training, model benchmarking/measurement, and model optimization
- Communicate results to scientific and business stakeholders
- Collaborate with other engineers, data scientists and business analysts.
- PhD in computer science, Physics, Statistics, applied Mathematics, or other quantitative/ Computational discipline or Master's with over 2 year of real time relevant experience in deep learning is a MUST.
- Expert knowledge in Convolutional Neural Networks (CNNs), RNNs, Transformers / self-attention, Graph Neural Networks (GNNs), Self-supervised learning, 3D data
- Fluency in high-level programming and scripting languages (R, Python)
- Extensive hands-on experience with one or more popular Deep Learning frameworks: TensorFlow, Keras, Caffe, Torch
- Proven experience with deploying models onto GPUs for production
- Having a passion for solving interesting and challenging big data problems
- Strong analytical skills
- Ability to work independently
- Ability to present data to a multidisciplinary audience in a clear and cohesive manner
- Excellent teamwork, time management and organizational skills
- Excellent verbal and written communication skills
- Data Analysis and Modeling using Python/Spark: 12 months (Essential)
- Time Series Analysis and Modeling: 6 months (Preferred)
- Distributed data processing using pyspark: 6 months (Preferred)
- Deep Learning on structured data: 12 months (Essential)
- Computer Vision/NLP: 24 months (Essential)
- MLOps: 6 months (Preferred)
Covid-19 update: Work from our office in Detroit, MI, remote work is not allowed.
About Soothsayer Analytics
Soothsayer Analytics is a Data Science company based out of Livonia, MI and offices in Columbus, OH and offshore in Hyderabad India. Advanced Analytics firm focused on Pattern Recognition and Unstructured Data. Our team actively works with, creates, and researches cutting edge Data Science techniques. We approach each problem individually and architect custom solutions. Our strength lies in our ability to build and productize proprietary algorithms and analytical tools. Advised by 10 industry experts whose domain knowledge we leverage to better architect industry specific solutions. Our delivery partners include 20+ Data Scientists with a combined 75 Patents and 300+ Publications. We collaborate with this brain trust to keep us abreast of state-of-the-art techniques and to help deliver world-class results. We utilize cutting edge Machine Learning and Statistical Techniques to extract Hidden Insights and Patterns from Complex, High Dimensional, and Unstructured Data. Our major clients include: DOW Chemicals, Ford, Visteon, D&B, Timken, US steels, Steelcase, Abercrombie, Express, Stanley Steemer, Whirlpool, AEP, NiSource, GEA, etc.
Soothsayer Analytics is an Equal Opportunity Employer and e-Verify Company.