The interim CEO asked me to determine costs as we considered beefing up our rendering pipeline by orders of magnitude.
Applying a KISS paradigm, I threw together an inital back-of-napkin estimate. That was good enough, but then the interim CEO needed a deeper level of understanding, so I worked with my Pakistani Technical Project Manager to build out a more comprehensive version of the model, taking parameters into account like the following:
Scale Factor
Model Throughput Per Week
Aspirational Efficency Factor
Cost Per Model
Number of Modelers Needed
Model Category
Results
Delivered cost model iteratively for a 3D modeling pipeline to support shop-the-room.
Leading-by-example to own and improve systems as sole ENGR having SRE/DevOps/Frontend/Backend experience.
Action
Mar 2019
Watching our AWS costs rise ~8% monthly…
Costs Rising
I learned about and subscribed to Reserved Instances to realize costs savings for our hosting spend:
Dec 2019
Though not leading to cost savings or revenue generation, part of my responsbilities have been database administration, jumping in when the production DB would spike like below, figuring out if a runaway process needed to be terminated, if a slow query was bringing it to its knees, if a cron job was introducing load, or whatever needed to be done to keep the site up.
Or when bots would crawl the site, bringing it down, necessitating an IP block:
Or when digging into the logs to find that a route was 500ing and had to be fixed:
Mar 2020
Using Cloudcraft, I diagrammed our AWS infrastructure, identifying and deleting 1000 unused SQS instances.
Also identified and deleted numerous unused RDS snapshots:
All changes led to a yet another 37% reduction in MoM AWS costs:
Results
Saved company 115% of my salary in 2019 through process improvements.
Take a half-baked v1.5 SOA and continue re-platforming to support scaled integration with corporate parent and its subsidiaries.
Action
Aug 2018
When I joined, it was with the idea that I would be instrumental in helping scale the Decorist relationship-creation paradigm leading to higher AOV to its corporate parent – Bed Bath and Beyond – and its subsidiaries: Buy Buy Baby and Cost Plus World Market.
I inherited the beginnings of a movement away from a Django/Angular/MySQL monolith and towards a Django/React/Postgres multi-tenant platform; a v1.5 of the application architecture. My predecessor had extracted some facets of the monolith to build out the initial stages of the SOA in partnership with another Bed Bath and Beyond subsidiary: One Kings Lane.
Although labeled a multi-tenant platform, it had been built quickly – effectively as a prototype – to service one tenant and was not truly extensible for other silos, though it did encompass solid SOA principles for reuse.
Dec 2018
As part of roadmapping for 2019, wanting to strategically position ‘scalability’ for the corporate org, I worked with my boss – SVP Prod/Tech – to detail a project plan to leverage the exisitng nascent SaaS paradigm for the next generation of the Decorist user experience.
We were sanctioned by the CEO for reduced scope on the roadmap; it would be the very first effort internally for moving towards a new platform to future-power the site. We focused on porting a back-office, manual task of determinging supply-side match and availability with demand-side need.
Wanting to keep a engineer engaged, I shepherded him and helped him understand the Cost-Benefit trade-offs we would need to make in order to meet the deadline, setting a course of re-using the User service for authentication, builidng out a new React-based UI, and integrating with the monolith.
Feb 2019
A few months later, we delivered the feature of supply-side Matching and Availability:
Shortly after launch, business priorities shifted towards a focus on a new eCom offering and we had to back-burner re-platforming efforts.
Dec 2019
Given guidance that 2020 might be the year we could re-visit re-platforming, I ideated around what was still lacking in v1.5 and what would need to be built out, creating a project plan and technical roadmap for the go-forward.
I worked with my boss again around planning and prioritization and we again provided guidance to CEO. Ultimately, we could not secure the necessary corporate integration buy-in for further progress and have had to defer any further progress given other priorities.
Results
Mentored Lead Engineer towards implementation of first feature for new SOA paradigm.
Given trends in 2018, AWS budget was projected to increase 92%.
Action
Worked with full-stack engineer on web team to tweak EC2 configs.
Iteratively introducing Reserved Instances to add to cost savings.
Worked with VR team to optimize the rendering pipeline.
As part of the VR work, we found that EC2 instances were running unneccessarily 24/7 so we adapted processes to more efficiently utilize those resources.
Contributors, as they are called, are the +5M people around the world who do work on CrowdFlower’s platform. The application that enables them to do work is one of the company’s heavily trafficked as well as most complicated – blending a Rails backend with MooTools, jQuery, and RequireJS in the frontend.
The application’s UX
…had largely stayed the same for the last five years. In Q1/2014, we decided to enhance it by making it more interactive and towards engaging our users more and conveying the just how much work there is in our system.
Working with the Product Manager and an external Designer, we came up with the following high-resolution mock
Because the application is so heavily used, we knew we couldn’t merely throw the switch on a new design overnight; both from a community management standpoint as well as application performance. Instead, we chose a strategy of introducing a first at the company: use of A/B Testing to determine a design that would perform as well as if not better than the original.
Our key metric for performance in that regard had to do with contributor’s performance after being exposed to the new UX, particular the messaging around our forthcoming gamification and introduction of Levels. In the beginning, we did not have the infrastructure to determine the value for that metric so we simply settled on ‘clicks’ as a (conversion) proxy to understand if the new design was having an impact.
Infrastructure
Without an A/B Testing framework in place, I needed to choose one. As requirements were not concrete for such, I did some due diligence in vetting several options, coming up with a review of A/B testing frameworks for Rails.
It became obvious that Vanity was best suited to our needs. (Since it doesn’t yet have the ability to throttle a percentage of the traffic receiving experiments, I augmented it with Flipper.)
Once that was in place, we could begin iterating on the design, knowing with confidence how we were impacting the user experience.
Server-side
We knew we wanted the experience to be snappy, but completely replacing the existing experience with a Rich Internet Application was far out of the scope for the first month, particularly as there were infrastructure changes to be made to retrofit the stack with A/B Testing. We decided to make progress iteratively over several sprints.
In our first test, we pitted the control (original) against a bare-bones implementation version of the high-resolution mock as the new design.
original
The new version out-performed control (in terms of clicks) 21.3% vs 20.3% (at 95% confidence) so I continued to iterate on the implementation, coming up with the following
To calculate the overall satisfaction by other contributors for a task (denoted by the stars) proved to be too inefficient in this iteration; it wound up losing.
Client-side
On the assumption that we needed to make the experience snappier in order to drive engagement, it was obvious that we would need to have more (and faster) interaction and therefore, an interactive client-side implementation.
As what was essentially a completely parallel product, leveraging only some of the infrastructure that the server-side rendition was utiziling, I begin to flesh out the following
Further refinement (an actual data) was necessary to get it looking more like the high-res mock (and like its server-side-rendered peer)
At this point, we implemented and integrated with our own homemade badging solution, beginning to display badges in the following iteration
The new version out-performed control (in terms of clicks) 21.3% vs 20.3% (at 95% confidence) so I continued to iterate on the implementation, coming up with the following
Testing the impact of particular messaging was also of interest, so we added a Guiders variation as well. At this time we also leveraged Google Analytics Events on the Guider buttons to track how the far the user got in our messaging.
Letting the experiments run a few days with sufficent traffic, we found that client-side-rendered version peformed no worse than the server-side-rendered version (23.9% vs 22.9%) and that having guiders also performed not significantly worse (23.1% vs 23.7%) so we decided to keep both.
By that time, the new version was out-performing control (the original design) 22.2% vs 20.7% (at 99% confidence) so a decision was made to move forward rolling out the new experience to 100% of contributors, doing some polishing (copy/styling) work before finally settling on the following
Results
Used A/B testing to upgrade company’s most highly-trafficked page (5+M views/month,) increasing user engagement by 5% and saving $2K/month (in Bunchball costs) by rolling own simple badging solution.
This was an enormous effort to overhaul a product whose UX had not been altered much in five years.
We took a piece-by-piece approach to swapping out components because of the complexity of the legacy behemoth. First, we refreshed the views in the legacy app, which involved changing styling in three different places (because the app had grown “organically” over the years, taking on three different styling paradigms styling was defined in custom stylesheets, in Less, and inline.)
In parallel, part of the team started building out the new peer Rails 3 app, the eventual destination for all views, complete with the company’s brand-new proprietary SSO solution (also built in parallel.) Finally, routing was updated to send all traffic to the Rails app.
Forming
Between August and September of 2013, we coalesced as a team under the project champion, the company’s CTO, and began formulating what the new UX should be and do.
Below is a screenshot of an example of the dashboard as seen by the end user (Merb, built in 2008)
Below is a screenshot of the progress of a microtask job, also as seen by the user (sensitive information redacted)
Norming
Between September and October of 2013, we cranked out the new experience.
Based on a design concept by the other F2E in the team, we began restyling low-risk interfaces of the system. The new design was not simply a reskin, but involved introducing a similar-yet-improved information architecture, an example of which can be seen below
Following are a few more example screenshots demonstrating the evolving look-and-feel
Configuration Panel
As we were tackling the UX, a backend engineer in a peer team was working in parallel to create a custom role-based SSO system that we would leverage for enforcing authentication and authorization in a new way for the company.
Shortly before the conference, a decision was made to go with a second design concept, not entirely different from the original, but a little more polished. A designer was requisitioned to provide the new design. From that point forward to product launch, we mostly fine-tuned the details.
The following screenshot demonstrates not only the new design but also the use of the new SSO solution, which can be seen where certain UI elements are disabled based on the user’s permissions
To QA the new experience, we ran it in alpha against production data repositories just prior to the conference.
Performing
After the launch, we maintained the product, adding features we had not been able to squeeze in.
Below is an example screenshot of how the final product shaped up
Results
Consolidated multiple styling paradigms for new UX ahead of company-sponsored conference.
The CrowdFlower platform is consumed via a number of microtasking sites. Each site registers and maintains its own users, but to better track unique identities across the CrowdFlower platform, we built a Single Page App in Ember.js to allow associating users across partner microtasking sites with one unique identifier in the CrowdFlower platform.
Results
Implemented a CRUD tool for managing users using Ember.js while iterating in conjunction with Product Manager as requirements changed.
The app is CrowdFlower’s most highly-trafficked app. It also happens to be one of the company’s most technically complex, given its history.
Its architecture is that of a Rails app, wrapping a Gem that extracted business logic from the company’s legacy (original) Merb app. The Gem contains all logic around rendering, styling, and providing interactivity for CML, the basis of abstracting microtasks in the platform.
The app was built (before my time) in order to bring a richer, more interactive experience to those doing microtasking work. When the original architect departed only weeks after I joined the company, maintenance and feature implementation fell to me.
Results
Supported site’s most highly-trafficked, revenue-generating UI (allowing for custom JS and CSS.)
