Jesrael Mani

- Fleet demand and allocation: Responsible for the models that decide where bikes and
batteries are placed in each city, including a real-time demand forecast in production since
2021, and for the measures used to judge whether those decisions worked.

- Optimisation of field operations: Responsible for the systems that tell field staff what to do
next – which bikes to attend to, in what order, an

- Credit scorecards were developed to identify customers likely to take a personal loan or credit
card, improving on the models then in use by 10–15 percentage points of Gini.

- Credit-report generation was migrated from legacy SAS to Python, and made ten times faster.

- Optical character recognition was developed to read identity documents and fill in online
application forms automatically, red

Metrics and modelling for banking and finance clients.

Worked as Data Science Trainee at Institute Of Analytics (USA).

Worked as Research Fellow in Theoretical Physics at The Institute of Mathematical Sciences, Chennai.

Six years of research in condensed-matter theory using advanced mathematics and Monte-Carlo simulation.

CGPA 9.4. Thesis: a profit-aware multi-agent reinforcement-learning approach for dynamic bike-sharing rebalancing, evaluated in a discrete-event simulator against a mixed-integer baseline.

Pursued a Master of Science in Physics at the Indian Institute of Technology Kanpur from 2004 to 2009.

Studied Physics at the Indian Institute of Technology Kanpur.