Brief : Beyond the analyst,scientist and ML engineer roles, what are the possible evolutions of roles surrounding data science and how we are seeing it develop across different projects in our agritech startup .

Takeaways

• Possible ways of structure a data collection, engineering and science team and pros and cons of them
• Walkthrough various low-code tools can be used to empowered roles that did not have access to or ability to build software earlier.

Outline

Introduction to Jiva

525 million smallholder farmers grow 70% of the world’s food, yet they’re among the poorest people on earth. We’re changing that.

Outline of the talk

Data science projects

• Credit models that predicts default probability of short term loans disbursed for financing crop procurement
• Crop doctor - plant disease detection model that can be used by farmers
• LLMs that enable the access of agronomy advise to indonesian farmers

Crop doctor

• data annotators (knowledge worker), active learning , users
• data quality filtering using reviewers
• focused labelling using active learning

Credit modelling

• data quality assessment
• data quality maintainers and processes to created to evaluate
• explainable models,compliance , Relationship managers who need to be model explainers

Data stewards / quality managers

• How sql and low code tools have empowered teams on the ground to create and maintain their own data processes and pros/cons around them