Sam Hardy | Thursday, November 15th, 2018
Problem: IT Recruitment Marketplace required an automated candidate matching service.
-Conducted multiple-user-workshops; establishing requirements for Artificial-Intelligence(AI)/Natural-Language-Processing(NLP)/Machine-Learning(ML) to automate user-journeys.
Conducted iterative (user-based-feedback) prototype-development based on high-risk backlog-items:
-Automated-CV-categorisation based on submitter-organisation/subject/type; sending through JIRA-workflows for processing - including conversion of action-items to JIRA-tasks for tracking through to completion.
-NLP using Tensorflow against candidate CVs; automating candidate matching to job description.
-NLP of web-scraped information from web-based-opportunities; leveraging Apache-NLP/JSoup.
-Smart-service-discovery using Hashicorp-Consul; facilitating multi-Cloud architecture.
-Data/file relationship-management using Neo4J Graph-database; facilitating rapid-search of high-volumes and access to related documents.
Outcome: Identified Beta-backlog, success criteria, budgets and team.