Engineer - Machine Learning

Recruiter
Bagboard
Location
London (Central)
Posted
12 Oct 2018
Closes
10 Nov 2018
Contract Type
Permanent
Hours
Full Time
*ABOUT BAGBOARD*


Bagboard is looking to change how advertising works, from the ground up. We empower consumers to be an active part of advertising and be rewarded for their engagement. At the same time, we’re giving brands a platform to engage in meaningful interactions with their audiences. We want to turn advertising into a two-way conversation and not a one-way broadcast.

At the heart of our platform is a consumer driven digital experience that actively encourages and rewards consumers to engage with our platform. The strength of our business model is that it incentivises consumers, advertisers and our retailer distribution network to not just personally gain, but also collaborate on utilising the value of our platform to solve wider social, environmental and economic problems facing us all. Advertising in the hands of the consumer

*ABOUT THE ROLE*

Bagboard is looking for an exceptional statistical ml/algorithm focussed engineer to join the team. Your mission will be to help the business utilise large volumes of data from our network to build solutions in two primary spaces:

* Logistics: Building living stochastic models of inventory availability for all our advertising distribution points.
* Advertising: Building a reactive, market driven media buying platform offering the best value for all media buyers.

These two areas are intimately linked, so we’re looking for someone who can truly own the overall engineering and mathematical spaces they encompass. You’ll be expected to lead our use of various internal and third-party data sets to build accurate models that will be at the heart of how the business operates. This role is at the heart of our most commercially critical processes, so would present massive growth opportunities for the right candidate.

Some examples of problems you’ll be expected to tackle:

* How should we create algorithms to employ the most appropriate forecasting model on differing inventory with differing amounts of historical data?
* How do we build a dynamic auction based commercial model on advertising inventory with lead times measured in month?
* How do we create a “fair” price for every piece of inventory before it has been exposed to the market?
* How do we tackle the problem of selling guaranteed amounts of inventory in the future on an advertising network whose capacity changes constantly?
* How do we build a pricing model that incentivises advertisers to purchase high value inventory, but also tries to ensure all inventory is sold at an acceptable rate?


*Requirements*

Essential

* Demonstrable commercial experience with the architecture, design and building of machine learning based applications and solutions.
* Strong Python capabilities, specifically with the major stats libraries (scipy, numpy, pandas, tensorflow, pytorch).
* Proven track record of building algorithms for real world environments at industry scale.
* Strong desire to work for a business built on the ideals of social, economic and environmental change.
* Database management experience: SQL/PostgreSQL

Nice to Have

* Worked in either the ad-tech, fin-tech or logistics optimisation sectors.
* Experience working with large scale first and third-party data sets.
* Familiarity with stochastic optimisation, linear programming and time series modelling practices.
* Experience working at a start-up, specifically with rapid scaling and growth.

Similar jobs

Similar jobs