Discovery phase and development of a web-based application for Spectrum

5 minutes read

Data is the new petrol but just working with data is not enough. To be more effective, you need AI algorithms to help you with the analysis of the data. Over 70% of IT vendors require a discovery phase to find a proper AI solution before starting the development.

About the client:

Spectrum delivers integrated marketing, communications, and media solutions hyper-focused on science. Fueled by the scientific method, the proprietary planning process of Spectrum allows to position the clients’ business challenges within the border landscape, leading to strategic, creative solutions across the marketing mix. Spectrum works for: biopharma, emerging biotech, established biotech, clinical trial recruitment, consumer science, public affairs, corp comms, health technology.

In 2020, Spectrum acquired VITech to make a discovery phase for finding out the proper AI solution for solving certain client’s needs.

Value delivered

As a result of the discovery phase, the client has received the discovery report with all required artifacts for developing the future solution: the software architecture design description, the software proof of concept document, the work breakdown structure, the team composition, the road map. Additionally, a web-based application was developed that provides faster and more reliable data analysis of client databases. It allows to facilitate and accelerate decision-making in the area of marketing.

Technologies used:

Java

JavaScript

Spring-boot

Python

MySQL

Kubernetes

GCP

JenkinsX

Tekton (VITech SDLC)

Discovery phase

Discovery phase and development of a web-based application for Spectrum

Business challenge

Data analytics by certain criteria has always been a challenge for marketing companies, including Spectrum. At one point, their data storage accumulated such a huge amount of information that it was quite challenging to derive meaningful insights from it within a reasonable timeframe. However, having data is far from being enough. What’s more important, is to make use of that accumulated data. This was the problem that Spectrum asked VITech to help with.

Why VITech?

By that moment, VITech had already developed Science puls, a solution capable of analyzing how frequently certain terms and expressions are used in the scientific literature over a period of time. This helped us to accumulate valuable experience working with the company’s complex request to the database.

We have started thinking about the new solution and working towards a Big Data-based calculation approach for healthcare years before it became a recognized industry need.

Discovery phase

Project description

Stage 1: Investigation

The first stage of our discovery phase consisted of the following components:

  • Stakeholders analysis to identify external and internal stakeholders for the future project
  • Interviews to define hypotheses
  • Validation of hypotheses using the import effort matrix technique which allows prioritization of hypotheses according to their importance and impact on the product
  • Development of a clickable interface
  • Usability testing to validate archived results by using a certain interface

The result of the first phase of the discovery was a validated list of hypotheses and a validated user interface. It addressed the basic client’s needs. However, we’ve identified a number of additional analytics-related issues that a web-based application needed to consider.

During the first phase, the following specialists were involved:

  • Project manager, who was responsible for managing the communication channels between Spectrum and VITech
  • User experience designer, who developed the user interface, conducted interviews and workshops for testing the clickable prototype
  • A business analyst who was engaged in the defining, validation, and prioritization of hypotheses
  • A software architect who was responsible for developing a technological vision for this product

Stage 2: Creating of web-based application

The new, enhanced web-based platform was specifically designed by the VITech team to provide quick answers to complex questions like:

  • How to communicate to the client that a certain PR campaign was effective and addressed the client’s needs?
  • How to evaluate the effectiveness of the campaign properly using the help of ML algorithms in the analysis of standard metrics such as impressions, click-through rate, cost per click, and others?
  • How to react more quickly by improving the negative KPIs on certain metrics?

To achieve such a result, VITech has built a web-based application with a clickable interface. The frontend developer and backend developer joined the team at this stage. Our SDLC allowed us to significantly reduce the development time at this stage. Instead of the usual 2 or 3 weeks, the initial development period took only 6 hours.

As the result of the second phase of the discovery, the technical view of the project was developed which includes:

  • Developing software architecture design for the next 3-5 years
  • Creating a proof of concept scenario which has to demonstrate the effectiveness of ML for solving client’s needs

Stage 3: Project management

The third phase of the discovery was concentrated on the project management issues. The following artifacts were developed: The software architecture design description, the software proof of concept document, the work breakdown structure, the team composition, the road map.

The resulting document of the final stage of the discovery phase was a discovery report consisting of the following parts:

  • Abstract (market analysis)
  • Introduction (investigation to discover clients needs)
  • Methods (for example, we used import effort matrix to validate hypotheses)
  • Results (the software architecture design description, the software proof of concept document, the work breakdown structure, the team composition, the road map, the list of validated hypotheses, the web-based application)
  • Conclusion (technical recommendations about useful or unuseful hypotheses, pros, and cons of each used methodology)

Results

VITech has developed the next-generation ML solution that provides faster and more reliable data analysis to accelerate decision-making in the area of marketing. It has integration capabilities to be expanded into other businesses.

The key advantages that VITech designed into the system architecture include:

  • User friendly functionality
  • Supporting complex queries and requesting data of different type/level
  • Supporting the custom combinations of rules and date ranges
  • Categorizing the population stratification by certain criteria
  • Faster and more reliable data analysis of client’s database
  • Facilitating and accelerating decision-making in the area of marketing

This would be an important step forward towards the market leadership in population analysis and gives a client capability to create requests for information, requests for proposal, requests for a quotation according to the specific needs.

The final list of artifacts received by the client includes:

  • List of all posited invalidated hypotheses
  • The prioritized list of validated hypotheses (validation board)
  • Clickable mockups
  • High-level software architecture design document
  • Proof of Concept
  • Web app (codebase)
  • Proof of Concept software requirements specifications
  • Demo guide
  • Work breakdown structure and team composition document
  • Discovery report

Our case studies

prev
next

Our case studies

Leave us a message

You’re in a good company: