Journey of Data Sciences Lab @ Simpl

Reading Time: 3 minutes

When we started in 2015, from our tiny office in Juhu (Mumbai), we pretty much had single person teams. We were seven people strong, including the founders, so one person per specialisation is not a bad statistic. As we scaled, so did our problems.

In the beginning, when you are starting a product company like ours, you do-not have any data to do a lot of “Fancy ML and AI”. What we needed was a system that could collect data and dashboards to visualise and derive insights from the collected data. So we started our search for a Data Engineer and a Data Analyst, thereby expanding the team size from one to three.

As we grew and started hiring more people to solve targeted, structured problems, we realised the need of clearly defining roles and responsibilities of Data Scientists who join us as part of the Data Sciences Lab.

Data Scientists by this time had become a very broad term and multiple roles came under this umbrella. We toyed with names like Business Analyst, Data Analyst, Machine Learning Engineer, etc for a while. All these names have got some baggage of their existence and a predefined meaning in different context and companies. So we needed something that suits our use cases, solves our problems and more importantly something we believe in.

Finally, after multiple debates, discussions and from wisdom of last 3 years, we come up with the below two tracks that one can grow into at Simpl as part of the DSL depending on their preferences-

  1. Data Scientist — Analytics & Inferences
  2. Data Scientist — Algorithms

Data Scientist — Analytics & Inferences

As the name suggests, this team takes care of the analytics and inferences need of the organisations. They work with multiple business stakeholders e.g. Products, Growth, Repayments etc to push data informed decision making in the organisation and solve business problems using data. They love turning enormous amount of data into compelling stories and visualisations that challenge conventional judgement.

A data Scientist (Analytics & Inferences) is a partially centralised arrangement, where at the end of the day everyone is part of one data sciences team, but they work closely with the feature team alongside engineers, designers and product managers.

They help democratize data-informed decision making across the organisation.

Data Scientist — Algorithms

This team builds intelligent data products. Folks in this team work and explore cutting edge algorithms to create smarter products.

They create data products that are highly available and scalable systems working with super low latency.

Irrespective of which track one choses, we expect them to be a sound coder. At DSL of Simpl, we are technology lovers, and we treat excel as the forbidden fruit.

While our business is adopting and responding to the changes, you have a chance to jump on this rocket-ship and be part of this amazing journey.

If what we’re building interests you, do mail in to with the subject line Blog reference — data.

We are growing rapidly and in the process trying to establish new tracks in DSL. You can find full JDs at the end of this blog.

We’re also hiring across roles — product, engineering, UX, UI. So do mail in. Just change the subject from data to whatever your role is.

[About the author — Raj is the Head of Data Sciences (Analytics & Inferences) at Simpl.

While he is not playing guitar or cooking some exquisite recipes, he helps businesses create intelligent products, currently @]

JD for DS (A&I)

You could find a DS(A&I) at Simpl

  1. Designing experiments for statistical A/B tests to help the products team decide on a feature
  2. Building automated dashboards and visualisations for data informed decisions
  3. Finding patterns for fraudulent behaviour
  4. Deciding True Business Metrics (We will soon write a blog post on this)
  5. Providing support during fundings to tell data backed stories
  6. Helping business understand user behaviour by doing descriptive and predictive analytics

What do we look for in a DS (A&I) at Simpl —

  1. You should have a decent coding skill in Python/R and rock-solid knowledge of SQL
  2. Structured and First Principle Thinking
  3. Sound knowledge of statistics
  4. Communication Skills
  5. Ability to abstract granularity of maths and talk business
  6. Ability to handle chaos and find structure while at it

JD for DS (Algorithms)

We have multiple problem areas that a DS-Algorithms focuses on —

  1. Improving underwriting using alternate data
  2. Creating solutions for flagging fraud in realtime
  3. Optimising portfolio risks
  4. Recommendation systems for smart communications

What do we look for in a DS (Algorithms) at Simpl —

  1. Sound understanding of data structures
  2. Good grasp of statistical learning techniques like Linear Regression, Logistic Regression, Decision Trees etc
  3. Knowledge and comfortable with tweaked implementation of performance metrics, cost functions etc
  4. Ability to write production ready codes. (If your code cannot leave a python notebook, you are not suited here)

Leave a Reply

Your email address will not be published. Required fields are marked *