[emeerson]

It would take Googlmart 5+ years to play catch up if not longer.

Google's too disorganized and they don't have any supply chain tech built up. Amazon has years of supply chain infrastructure (both software, physical systems and organization) including supply/demand forecasting models and warehouse operational investment (Kiva systems, etc...).

The only differentiation that Google adds to Walmart is: - Top-talent engineering resources & brand which attracts talent to work on these problems - Layers of developer tooling and services that remove the "systems at scale" development curve of startups and smaller companies.

Now if Alibaba teamed up with Walmart... (read: acquired).

[throwaway6498]

In Bangalore, companies like Google, Amazon, Walmart Labs, Flipkart, Directi, and LinkedIn all attract the same talent pool.

All of these companies can pay either as much as Google or more than Google to attract the topmost talent pool. In fact, many of the engineers in these companies frequently hop from one company to another between these. Of these, Amazon has the highest attrition rate due to stressful working conditions and they are sooner to leave to another company than the others.

Walmart Labs in Bangalore has a talent pool that is running a million node cluster based on OpenStack completely in house with zero dependency on something like AWS (of course!) or GCP. That's quite a feat and I think it will help Flipkart a great deal if Walmart Labs shares some of this incredible talent and resources they have built within their company.

Disclaimer: I work for Flipkart.

[gbrayut]

FYI: http://oneops.com is the tech stack used for "Application Lifecycle Management of Cloud Based Workloads" at Walmart / Sam's Club / Etc. It was aquired in 2013 but remains open source (not sure who else is using it though). At a high level it is built around Chef solo, and allows you to define application "assemblies" that can then be deployed to various clouds (on prem or public PaaS). Being cloud agnostic means it can shift workloads between different cloud providers with only minor tweaks to the applications.

It works well for a large company with many moving parts and a need for Enterprise wide policies/quotas/etc, but has some rough edges and leaky abstractions. Kubernetes is a more mature and widely used system that has many of the same features, just based around containers instead of VMs.

Disclosure: I work for Walmart Labs on our CDN / Edge Platform team.