Data Scientist

About Geminus

Geminus is an industrial process centric digital twin product that delivers self-optimized design, predictive operational intelligence and asset performance enhancement to operators and engineering service providers. The product leverages physics-constrained artificial intelligence to deliver high-fidelity operations and maintenance intelligence and to continuously optimize the designs of assets & processes. Geminus integrates with operational data through SCADA systems, historian software, sensor gateways and enterprise asset management to trigger predictive modeling for the end-to-end industrial process. This generates high-fidelity operational intelligence and design recommendations to prevent multi-step failures, adverse shifts in yield, structural process faults and degraded remaining useful lifetime.

 

The company is based in Palo Alto and venture funded by The Hive. Its co-founder and Chief Scientist, Karthik Duraisamy, has over a decade of rich experience in physics-based modeling and machine learning through both fundamental research at Stanford University & the University of Michigan and industry collaborations. The fast-growing R&D team combines experiences across physics modeling, AI, industrial domains and data operations. The product is in pilot deployments in oil & gas and hi-tech manufacturing verticals. The product is also being made available as a module in leading engineering simulation platforms.

About the Role

The Data Scientist will drive the design and development of key artificial intelligence (AI) components of the platform that integrate with its multi fidelity physics-based modeling capabilities. The spectrum of applications of AI includes black-box IoT modeling,consolidated error modeling across multiple physics-based models, supervised & unsupervised training of parameters & data-driven functions in hybrid physics-based models and meta-learning for performance boosting.

Responsiblities

As a Data Scientist of a fast-growing startup, the successful candidate will be leading the development of key aspects of Geminus’ product:

  • Model training, model analytics and serving using state-of-art TensorFlow based orchestration techniques
  • Selecting features, building & optimizing predictive accuracy and error modeling using a combination of machine learning and deep learning techniques
  • Design and development of the AI components of the data augmented physics-based model training from batches of operational data across multiple heterogeneous physics-based models
  • Design and development of the AI-components involved in model analytics, selection and servicing for delivering optimal design & operational intelligence
  • Enhancing data collection procedures and data quality management from both numerically simulated and operational data sources
  • Processing, cleansing, and verifying the integrity of data used for analysis
  • Creating automated anomaly detection systems and constant tracking of its performance

About You

The successful candidate will have experience in working in innovative projects with fast-paced delivery schedules in startups & large enterprises:

  • Proven track record of analyzing large-scale complex data sets, modeling and machine learning algorithms
  • Excellent understanding of machine learning techniques and algorithms, such as k-NN, Naive Bayes, SVM, Decision Forests, etc.
  • Experience with common data science toolkits, such as R, Python (NumPy, SciPy, Pandas), Matlab etc.
  • Experience in deep learning frameworks (e.g., Tensorflow, MxNet), and Large-scale optimization preferred.
  • Good applied statistics skills, such as distributions, statistical testing, regression, etc.
  • 3-5 years of experience in Applied Machine learning
  • Educational background or prior experience in engineering physics and/or simulation is desirable

 

Geminus is an affirmative action employer and welcomes candidates who will contribute to the diversity of our team.

Please send your resumes to jobs@geminus.ai