Data Scientist (Algorithms)

Analytics and Business Intelligence

About the company

The analytics and BI team at Delhivery is a “business first data team”. We are part of both the technology and business teams, and our mandate is to identify and solve high-impact business problems using data and quantitative methods. To put it simply, everyone in the team is good at both understanding business and playing with data, and that is what we expect out of future employees as well.

We deliver more than 10 packages each second, and each package gets scanned, on average, 15 times. Conservatively, that is 13 million records a day from package scans alone. The scale and complexity of the business, and data, means that there is a massive opportunity for our team to have an impact on the firm’s finances and operations.

A large part of the work consists of understanding and manipulating data, finding patterns, linking it to existing and prospective business problems, and telling stories linking them. Along the way, we build some models, write some “smart” reports and build the odd dashboard. Occasionally we might put a model in production in collaboration with the tech and product teams.

Each person in the team takes end-to-end ownership of an entire problem statement – from talking to the concerned teams and defining it, gathering and putting together the data, analysing the data, building models and finally telling a story in a way that can be easily implemented by business.

Culturally, we are ambitious (in terms of the work we take on), experimental (always looking for new and improved ways of doing things) and work at a fast pace. The problems we take on are necessarily ambiguous and vague, and each member of the team is expected to take end-to-end responsibility, and make decisions independently.

About the team

The analytics and BI team at Delhivery is a “business first data team”. We are part of both the technology and business teams, and our mandate is to identify and solve high-impact business problems using data and quantitative methods. To put it simply, everyone in the team is good at both understanding business and playing with data, and that is what we expect out of future employees as well.

We deliver more than 10 packages each second, and each package gets scanned, on average, 15 times. Conservatively, that is 13 million records a day from package scans alone. The scale and complexity of the business, and data, means that there is a massive opportunity for our team to have an impact on the firm’s finances and operations.

A large part of the work consists of understanding and manipulating data, finding patterns, linking it to existing and prospective business problems, and telling stories linking them. Along the way, we build some models, write some “smart” reports and build the odd dashboard. Occasionally we might put a model in production in collaboration with the tech and product teams.

Each person in the team takes end-to-end ownership of an entire problem statement – from talking to the concerned teams and defining it, gathering and putting together the data, analysing the data, building models and finally telling a story in a way that can be easily implemented by business.

Culturally, we are ambitious (in terms of the work we take on), experimental (always looking for new and improved ways of doing things) and work at a fast pace. The problems we take on are necessarily ambiguous and vague, and each member of the team is expected to take end-to-end responsibility, and make decisions independently.

About the role

We need crazy (no exaggeration) algorithm designers to “perform gymnastics” with the firm’s data, draw analogies and find patterns, and tell stories that have not yet been told, and which will help the firm solve pressing business problems.

A very small sample of the problems you might work on:

• Design and implement smart dashboards and reports that tell the whole story by themselves, without relying on the user to drill down or make choices. For example, can the CEO get a daily automated report that talks about everything significant that happened in the business the previous day?

• Recommend how to best utilise our trucks for in-city deliveries in order to reduce costs. The approaches we have come up with so far for this include some combination of computational geometry, genetic algorithms and operations research.

• We do a large number of our last mile deliveries using independent contractors. How do we price our arrangements with them?

We need fundamentally curious people who are happy to try new (and potentially unusual) approaches to solving the problems. We also need “algorithmic generalists” – people who are familiar with a very wide range of algorithmic approaches, and are always willing to learn and try out new things.

Broad responsibilities:

• Understand business problems and translate into computational / data problems. And then convert the solutions back to the data domain

• Develop models, algorithms and heuristics to solve business problems in an efficient and time-bound manner

• Develop logical metrics to evaluate and monitor the business

• Build impactful and automated representations / visualisations of underlying data and key metrics in order to enable the business to make key decisions

The role reports directly to Karthik Shashidhar, Senior Vice President, Analytics and Business Intelligence.

Required Qualifications

We are looking for people who are excited by data, and excited by business, and excited in bringing the two together to create high impact. We are a fast-paced team, and expect quick turnarounds. Most of the problems we work on are ambiguous, and resolving ambiguity is a key part of the role.

Key requirements

• Bachelors/Masters degree in Computer Science and Engineering / Mathematics / Statistics from a top-tier institution (either in India or abroad)

• Strong at algorithms, classic artificial intelligence (not machine learning) and heuristics

• Very good at mathematics and logic

• Willingness to learn and use new technologies and programming languages

• Ability to work with “big data”

• Strong written, oral and visual communication skills

• Common sense, and an appreciation of common logical fallacies

Knowledge of R is a massive bonus.

For this role, we are hiring across experience levels. We pay market salaries, including ESOPs.