Technology search service

Context

role: UX/UI designer

role: UX/UI designer

role: UX/UI designer

duration: 9 months

duration: 9 months

duration: 9 months

A service for searching technologies for Gazprom-NTC and their partner companies. Gapzrom-NTC is a scientific subdivision of the largest oil producing company in Russia, whose activities are focused on the development of technologies in resource extraction.

The search is a mixture of GPT and filters of a regular online shop. Companies can offer their developed technologies and declare their need for a technology. A representative of a company that needs a technology can set up the usual filters like in an online shop, but for an in-depth and precise search they can use ML search.

Various resources can be used to make an advanced ML search work: documents of different types related to the technology, articles about the problem the company is trying to solve, terms of reference, and the ability to freely describe what is needed. This increases the chances of finding what is required in an uncertain problem statement.

What was my task

Gazprom-NTC is the scientific division of a major gas and oil producing company, which co-operates with many partner companies that provide it with their services as various technologies. Due to the lack of a single service where companies could exchange and search for the necessary technologies for their needs, this greatly slowed down processes and increased project implementation costs.

What I have done

  • Designed the desktop web interface of the inter-company technology exchange platform

  • Worked with the head of ML development and stakeholder Gazprom-NTC to clarify the final vision of the platform, identify the problem, analyse competitors and create a user flow

  • Worked with a ready-made Consta design-system created by other designers

  • Created an interactive prototype of the main flow

User problem

Due to the lack of a unified platform, information about available solutions was lost or unavailable. Many existing technologies were not utilised effectively and were not scalable to other industries. These problems caused negative consequences for Gazprom's key stakeholders and partner companies. Project managers faced delays in deadlines and reduced team productivity. Engineers experienced difficulties in accessing already developed solutions and faced duplicate tasks. Contracted technology supplier companies faced lengthy approval procedures, unclear requests.

Understanding UX and user flow

In the first iteration of the built user flow the user searched for technologies only by setting up the usual filters. If the user did not get a result, the user had to create a card for the technology and then ML search was connected and offered a result. Based on the results of the tests, this search process was long and inefficient. Users got irrelevant results by superficial filters. And in order to correct the ML search result, it was necessary to edit the technology card itself.

First user flow improvement

At this stage it was decided to put ML search in a different section. Users need to go through all the steps to get ML search results. The usual filters remained available on the general list of technologies. But in the end it made the search process less intuitive and complicated navigation. Many actions were required before getting a result.

Second user flow improvement

In the last iteration, the ML search was combined with the main search process, together with the regular filters. Users could first adjust the regular filters and then get improved results by filling in the ML search fields. The ability to adjust the regular and ML filters was also facilitated, receiving changes in the search results immediately.

However, while improving usability, users sometimes found it difficult to work with the combined filters. In order not to scare the user with technological terms, ML search was named ‘advanced filters’ and was accompanied by tooltips in the form of tooltips, how this or that filter helps in the search.

The UI

Working with Gazprom's design system, Consta, was an important part of the project as it ensured consistency with the brand and the company's other products. Having learnt Consta, I applied its elements in mock-ups and prototypes, which accelerated the creation of the interface and will simplify the work of developers.

List of technologies

Users can set up simple filters for quick searches and fill in the advanced filter fields to run ML searches

List of technologies with ML search

In the text field add information about the technology and expectations from the result of its work in free form. Also any information that will help ML filters to work is included here. Users are given some ideas of what can be written in the field, in the hint.

There are many industries for which technologies can be used and some of them can be applied to several industries at once, in different industries. The text box will assist in the search by substituting results as you enter characters.

Users can select pre-loaded documents from the Gazprom database so that users don't have to wait for downloads or create duplicates in the repository. If there are no technology-related documents, they can upload new ones.

Creating a request card when there is no search result

If the user does not find the technology they are looking for, they are prompted to create a request for that technology. In order not to force to fill most of the fields, the platform will pre-fill the fields from the both filled filters. It will also draw the user's attention to them through colour, offering to check if they are filled in correctly.

Problems and solutions

This was a challenge for me as this is my first project as a designer. There were a lot of questions about design processes, which I knew almost nothing about before. There were a lot of mistakes made and a lot of resources learnt while building the platform.

Initially, there was no understanding of how the platform should work and where ML search would be included. In the end, the solution was found after a number of calls and discussions with the stakeholder and ML team leader, several user flows created and sketches drawn.

Unfortunately, the development of the platform was hit by the crisis caused by Covid-19 and no perspectives for funding were found. The project had to be closed.