Client: Payback - leading loyalty program.

Problem: Three separate content-management systems needed to be merged into one and integrated with a transactional system developed abroad, in the company's HQ. The previous solution didn't scale to Payback’s massive amount of users, so a fresh technological approach was necessary.

Challenges: Replace multiple CMS solutions with one, demanding high scalability, based on a new technology stack. At the same time, don’t exceed budget constraints, and deal with a constantly changing environment.

Approach: Operational and budget limitations meant that it was not possible to build the whole solution at once. Thus the system was replaced piece by piece during regular development.

Outcome: Three separate CMS panels were merged into one and integrated with the external transactional system. Direct costs took about 10% of the relevant budget — the other 90% was covered by current development costs. The entire process of development and exchange was performed “on the run” without any system downtime.

Client: Leading fashion retailer (NDA)

Problem: Indicate loyal customer segments suitable for marketing purposes.

Challenges: Normalize the annual raw transactional data, marking out meaningful customer segments and archetypes.

Approach: First, error-packed data fields were normalized and unified, reducing the number of products and categories. Next, attribute extraction reduced the huge database into a limited list of products according to analyzed attributes. Then customer archetypes and representations were designated and clustered into meaningful segments.

Outcome: Over the course of two weeks we composed and presented meaningful customer segments, each described with their financial value, demographic analysis, buyers’ preferences, and archetypes, together with the identification methods we used. Individual customer IDs were assigned to their relevant segments.

Client: Automater.com - sales platform for digital products.

Problem: Provide the 15,000 stores hosted on the Automater platform with quality information about online users’ behavior and sales value.

Challenges: Ensure continuous delivery of highly accurate data to a significant amount of separate accounts in real time.

Approach: To achieve excellent data accuracy and real-time data processing, we connected JavaScript data collection with our proprietary Retailic Business Intelligence platform (based on Apache Storm software). We designed an API to send process information and individual data to every Automater account, simultaneously.

Outcome: During seven working days we adjusted the Retailic BI tool to Automater’s needs and delivered a customized API that could present business information to thousands of stores in real time.

Client: Telecommunication company (NDA)

Problem: Identification of IVR nodes where customers disconnect without achieving their goal. How does IVR structure change before and after modifications?

Challenges: Modeling the customer decision tree for actual IVR structure.

Approach: The complex nature of the given IVR forced projecting its structure into a decision graph, instead of a decision tree. After we normalized the primary log data, an application was prepared that contained individual node traffic information. The application featured an interactive, intuitive, and configurable graph of the user’s journey.

Outcome: In two months Retailic prepared the interactive application, allowing managers to easily explore the decision graph of IVR callers.

Client: Fablify.com - marketing efficiency tool

Problem: Identify the marketing campaigns that generate the highest income and boost buyer-acquisition rate, on a platform used by more than 150,000 entrepreneurs.

Challenges: Attribute every online purchase to a specific marketing campaign, nail down its source or type, identify users as a new or returning customers, and display all of this information on a massive scale.

Approach: We elaborated a user-tracking method that connects the parameters used in tagged marketing URLs with users’ visits to the stores, as well as their purchase time and value.

Outcome: After two months we had prepared an online application that was fully tested and integrated with Shopify.com (a leading platform for smallscale e-commerce).

Duties:
Creating dynamic frontend applications.

Requirements:
- Experience working with HTML, CSS, and JavaScript.
- Knowledge of the AngularJS framework
- Knowledge of React.js is preferred
- Good communication and interpersonal skills
- A passion for programming
- A focused, ambitious, and solution-driven approach

Code quality is the most important. Show us a repository with nice code in ClojureScript using React.js, that communicates with servers through Sente, and we hire immediately.

If you have self-discipline, remote working is not a problem.

Are you intrigued? Send your resume, GitHub link or example of code to: work@retailic.com

Duties
Creating and development backend applications.

Mandatory Requirements
- Comfortable with one of following programming languages: Clojure (or other Lisp), Scala, Ruby, Python, F#/OCaml or other functional language
- Good knowledge of SQL and relational databases
- Knowledge of concurrent programming
- Familiarity with web application development
- Familiarity with client-server communication, esp. with HTTP protocol
- Comfortable with Unix-like systems
- Comfortable with version control system (git preferred)
- Working knowledge of English language

Preferred skills
- Experience with web application development
- Experience with TDD methodology
- Working knowledge of Java and JVM
- Experience with PostgreSQL
- Experience with "NoSQL" databases (Redis especially)

We mainly use Clojure, but we do not require it at the beginning: knowledge of any language function enables you to easily learn Clojure.

If you have self-discipline, there is no problem with remote working.

Interested? Eager to join? Send us your resume, GitHub link or example of code to: work@retailic.com