Dmitry Grenev

VK Data Portal

Internal data/AI products and shared UI foundation for analytics workflows at VK AI / Data Office.

The main design work was to turn dashboard behavior into a controlled product system: shared controls, chart grammar, admin flows, and reusable VK Sky components.

NDA-safe case. Confidential metrics, internal datasets, and business-sensitive details are intentionally omitted.

Home →CV →

Case

1Overview2Facts3Home

Story

1One workspace for reading and controlling reports.2Reports became a structured product space.3Settings stayed inside the analytical context.4Dense reports needed a focused reading mode.5Maintenance needed its own interface.6Reporting needed ownership and distribution.7Charts became reusable product modules.8Benchmarks lived inside the graph.

Role

Product Designer

Period

2023-2024

Context

VK AI / Data Office

Ownership

Sole designer

Domain

Internal analytics · Data-heavy UI · Design systems

Scope

UX flows · dashboard system · specs · components

Design

Dima Grenev

Team

VK internal team

1

One workspace for reading and controlling reports.

Navigation, report content, and contextual controls stay together instead of becoming separate dashboard fragments.

The main shell keeps analysis, navigation, and report actions in one place.
2

Reports became a structured product space.

NPS, Mediascope, and CM Reports share one model, so moving between reports does not mean relearning the interface.

A shared navigation model makes recurring reports easier to scan and maintain.
3

Settings stayed inside the analytical context.

Workspace preferences open over the report, so the user adjusts the system without losing their place.

System settings are treated as part of the portal shell.
4

Dense reports needed a focused reading mode.

Panels collapse when the report needs priority, while navigation and controls remain recoverable.

The focused feed gives the report more space without making the interface feel detached.
5

Maintenance needed its own interface.

Product settings, priorities, keys, and mappings became editable objects inside the platform.

Admin screens make the reporting product maintainable.
6

Reporting needed ownership and distribution.

Mailing and notification controls made recurring reports part of the product workflow.

The portal includes the people layer around analytics.
7

Charts became reusable product modules.

Legends, periods, hovers, goals, deltas, and actions became repeatable behavior across report families.

The chart module defines repeatable behavior for data-heavy UI.
8

Benchmarks lived inside the graph.

Competitor deltas and benchmark states appear at the moment of reading, not as detached commentary.

Benchmark context becomes part of the interface, not external commentary.
9

The work was to make analytics behave like a product.

The design value was in the system: shared report structure, reusable chart behavior, admin workflows, and enough control for internal teams to maintain analytics without rebuilding the interface each time.