Requests are all you need - CPU Limits and Throttling in Kubernetes Background At Numerator, the Deep Learning Team hosts numerous models as a collections microservices on Kubernetes. We have a variety of building blocks in our architecture that allow us to build more complicated pipelines and processing steps. At the core, we have a fairly large number of pods that are
Accessibility Inclusion, Accessibility, and Internal Documentation Introduction As any company grows, it requires documentation to keep track of processes, architecture, cross-team functionality, and on and on and on. To keep things organized and moving smoothly, we need to employ several kinds of writing, from administrative to conceptual to procedural. What is the difference? * Administrative: reports, contact
Engineering Accessibility & Insights Introduction > Web accessibility, or eAccessibility, is the inclusive practice of ensuring there are no barriers that prevent interaction with, or access to, websites on the World Wide Web by people with physical disabilities, situational disabilities, and socioeconomic restrictions on bandwidth and speed. - Wikipedia [https://en.wikipedia.org/wiki/
Solr Scaling Apache SolrCloud at Numerator Scaling Apache SolrCloud at Numerator, by reindexing to Time Routed Aliasas.
Joining Numerator It was a bright sunny day when I received a call from a talent scouting partner telling me about a leadership position with Numerator. Numerator was mulling over the idea of establishing a brand new workspace in Pune (MH), India. Although I was not actively looking for a job change,
Microservice Dev Environments Scene At a boundary between early morning and a late evening in February, two things were rapidly becoming shorter in my world: the days and Q1 project deadlines. To meet one of these deadlines I asked the survey team to create two new services. This was being done to provide
Cascading failures - a travel adventure On a not so recent trip to India, my colleague Kevin and I were subjected to some pretty terrible airport travel challenges. I hate to say but Newark is the setting here. I will walk through 10+ things that happened over 4.5 hours which looked shockingly like an application
Numerator at Pycon Canada 2019 Pycon Canada is the biggest Python event in Canada every year. Numerator has significantly benefited from the Python ecosystem as Python powers much of the work we do. This year we sent six developers from Numerator’s Ottawa office to Toronto to attend PyCon Canada 2019, it was a chance
AWS ReInvent 2019: A Recap Numerator provides education and training resources to our R&D engineering staff so they can continuously evolve and adapt to our business needs against the ever-changing software development landscape. Often this translates into tech conferences, and the DevOps teams' conference of choice is AWS ReInvent. AWS ReInvent is
Impostor Free Zone: Data Science at Numerator My name is Alex Kreig, Head of Data Science at Numerator, and I have a secret to tell you: most of the time I feel like I don’t know enough. While “enough” is admittedly an arbitrary measure of competency, the chasm between what I think I’m capable of
how-to Failure as a Service April 2017, about 2 months after I joined the Insights Platform [https://www.numerator.com/solutions/people-and-shopper-insights] team in San Francisco, we noticed an all-too-familiar issue for a startup: we were moving so quickly (at least 1 release per week) that we weren’t thinking about edge cases and scale.
On Demand QA Environments The Problem Let's just lead with this. The Engineering team at Numerator will double in the next year... An engineering team generally doesn't go from 20 - 40 or 40 - 80 without some demand from customers. This demand from customers requires more engineers to build
Hackdays at Numerator This past week saw the conclusion to our most recent company-wide Hack Day at Numerator. Each quarter we set aside 2 days for anyone to work in teams, or solo, on whatever they want uninterrupted. These projects can be anything from: an engineering hack building some new tool, a data
Auto-Scaling Vertica Clusters With consumer purchase data, business questions are only limited by imagination. * What is the purchase frequency for Pepsi among Hispanics in the Northeast? * Share of wallet Amazon has among regular K-Cup / Coffee Pod buyers? * Bang Energy Drink, is it stealing sales from Monster or Red Bull? This level of flexibility
Large Scale Migration to Amazon Aurora Recently, the platform team at InfoScout completed a large infrastructure migration to swap our primary transactional database from AWS RDS MySQL to AWS Aurora. The project was a success on many fronts and we were happy to observe material performance gains across the services that power our consumer mobile apps.
Receipt Store Detection via Tensorflow Inception Model The Motivation InfoScout’s data analytics and business intelligence platform is built on top of receipts. A lot of receipts... Receipts today come in many different shapes, sizes, and vary from retailer to retailer. Most differences are fairly simple, such as length, or possibly having a coupon or two which
HPE Vertica at InfoScout Our dataset at InfoScout is unique. For one, it's large - now the largest purchase panel in the world providing item-level data across all retailers. Secondly, all data is tied to unique households which differentiates it from POS datasets widely available. To extract maximum value from this data,
Bananas, Hananas, Jananas: A Tale of Typos The core of InfoScout’s business intelligence platform is built on top of receipts. Hundreds of thousands of receipt images are uploaded by shoppers from all over the country, every single day. The quality and resolution of these images vary wildly: they could be rotated, shot in low light conditions,
