Indigo is a company dedicated to harnessing nature to help farmers sustainably feed the planet. With a vision of creating a world where farming is an economically desirable and accessible profession, Indigo works alongside its growers to apply natural approaches, grow healthy food for all, and conserve resources for future generations. Utilizing beneficial plant microbes to improve crop health and productivity, Indigo’s portfolio is focused on cotton, wheat, rice, corn, and soybeans. The company is headquartered in Boston, MA, with commercial and customer service based in Memphis, TN. www.indigoag.com

The Machine Learning Scientist will drive innovation throughout R+D by uniting laboratory, greenhouse, field, Indigo Partners, and commercial data into actionable insights. As a member of the machine learning team, they will design, denoise, and analyze experiments from the laboratory to commercial fields to rapidly and repeatably identify beneficial microbes. They will combine image analysis and machine learning to allow the remote collection of phenotypic data from experiments at the plant to the field level. The ideal candidate will be an excellent communicator, who enjoys working with scientists and growers, quantifying their work and extracting as much information as possible using tools of modern machine learning. The machine learning scientist will have a unique opportunity to drive innovation and discoveries throughout Indigo. We desire a quick and flexible thinker, who is always looking for the next way to connect disparate data and to implement the best recent technical advances to advance Indigo’s R+D pipeline.

 

Outcomes:

  • Develop strategies to connect data at all stages of the R+D pipeline and use machine learning to speed and improve the quality of results
  • Use computer vision and neural networks to extract large-scale phenotypic information from laboratory, greenhouse, and mass field experiments.
  • Build causal machine learning models of large-scale Indigo Partners and Commercial field data

 

Competencies:

  • Using machine learning to gain scientific insight into complex datasets, in both supervised and unsupervised/exploratory contexts
  • Distilling images into relevant structured data using modern machine learning tools such as Convolutional Neural Networks
  • Statistical and probabilistic modelling of data, especially in the biological sciences.
  • Design of scientific experiments, e.g. power analyses, decomposing sources of variation
  • Writing reusable, comprehensible software that can be productionized 
  • Presenting technically sophisticated analyses to audiences at disparate levels of sophistication
  • Agile with the ability to deliver in a fast-paced environment
  • Ability to understand needs of customers, from laboratory scientists to growers, and experience working directly with stakeholders to implement exactly what the customer needs
  • Desire to teach quantitative and statistical skills and continually increase the level of sophistication in experimental design and analysis
  • Strong desire to continue learning, identify new techniques and technologies, and rapidly implement them to keep Indigo at the cutting edge (e.g. reading bioRxiv/arXiv daily and testing new tools)
  • Team player, excellent communication skills
  • Passion for Indigo and our core values

 

Qualifications:

  • PhD in physics, engineering, computational biology, computer science, statistics, or other quantitative discipline –or– Masters degree with 5+ years experience
  • At least 3-5 years' experience with programming in Python or another scripting language
  • Fluency in statistics and data analysis and an ability to pair technique and problem
  • Experience in major machine learning software tools such as scikit-learn and tensorflow
  • Experience with probabilistic programming tools such as STAN
  • Experience working with and analyzing imagery to extract patterns of interest
  • Understanding of and experience working with biological data

 

Apply for this Job
* Required
File   X
File   X


Share this job: