Konstantin Zudov

Saint Petersburg

Open to relocation and remote work

Experience

  1. Founder at Neurica

    2026 — present

    My own studio: we take on products for clients as a contractor and carry them from research to working code.

    • Inosago: online car insurance with a customer side, a sales workspace and an admin panel, 64 screens for four roles.
    • Monogrid: portfolio sites and client albums for photographers, designed and built to production.
  2. UX/UI designer, pilot project at Konsol.Pro

    2026

    Pilot in the design department: a multi-agent system that turns a product manager’s PRD into screens built from the company design system, without designers. Agents check the PRD for gaps and review every result.

    • 3 products and 320+ screens built entirely from design-system components.
    • Rebuilt the Lovable pilots from existing components and cut their technical debt.
  3. Design Lead, HR-tech at AlfaStrakhovanie

    2024–2026

    Internal ecosystem: org structure, business travel, bonuses, payroll and corporate healthcare. Contracted through Intelsy.

    • Made prototyping 3× faster with AI tools (Figma Make) and a leaner approval flow.
    • Built the design system with tokens and semantic versioning of components.
  4. Creative Director / DesignOps at Intelsy

    2023–2026

    Led the design department across internal and client products: strategy, hiring, quality control, handoff rules and regular crits.

    • Grew the design department’s profitability by 11% a year.
    • Launched 3 internal AI services: a multi-agent estimator for incoming projects (Neuro Goose), hiring automation and social media automation.
  5. Design Lead, B2B fintech at Moscow Exchange

    2023–2024

    Professional B2B bookbuilding platform for collecting orders and allocating securities in placements. Contracted through Intelsy.

    • Rebuilt the missing requirements through stakeholder interviews and aligned them with business, IT and compliance.
    • Designed roles, permissions and edge cases; delivered interactive prototypes and specs.
  6. Senior Product Designer, e-mobility at Wize, UAE

    2022–2023

    Ecosystem for commercial electric vehicles in Dubai: a driver app and a battery monitoring platform, 50+ back-office and 15 app screens under one set of interface standards.

  7. Senior Product Designer, social / gamification at Overclout, Thailand

    2021–2022

    Social network with gamification for the Thai market: feeds, profiles, game mechanics and UX localisation, within Flutter’s constraints.

  8. Art Director at Newmen

    2019–2021

    Digital client projects.

  9. Art Director at STEM

    2017–2019

    Visual identity systems for corporate events.

Education

  • 2008 Ural State University of Architecture and Art Faculty of Design, Yekaterinburg
  • 2024 Systems management of organisations (MBA) Simbirsk-LINK
  • 2024 Innovation through Design: Think, Make, Break, Repeat The University of Sydney, Coursera
  • 2024 Introduction to TRIZ Sergey Gorpinenko
  • 2023 Google UX Design, 5 courses Google, Coursera

Languages

  • Russian — native
  • English — B1
  • Spanish — A1

Skills

  • UX
  • UI
  • Design systems
  • Usability
  • Figma
  • Framer
  • Webflow
  • Pixso
  • Spline
  • Agile
  • MVP
  • LLM
  • Multi-agent systems
  • Prompt engineering
  • Claude Code
  • Cursor
  • Lovable
  • Bolt.new
  • ComfyUI

Contacts

Goose

An AI service that turns a project brief into an estimate: hours, cost and timeline, task by task.

Client
Intelsy, internal product
Dates
2025
Role
Product design
Platforms
Web
Category
  • AI
  • UX/UI

Overview

Intelsy.Оценка, codenamed Neuro Goose, is Intelsy’s internal service for estimating incoming projects. A manager describes the project in a sentence or attaches a brief, and the AI breaks it into analytics, design, development, testing and management, estimates every task in hours with a risk buffer and prices it at the team’s own rates.

The estimate stays editable: hours and rates change in place, tasks are added or removed, and the totals update at once. Behind it sits a control centre where the team keeps its rates, coefficients, project templates and the prompts the model works from.

Problem

  • Estimating a project by hand takes a lead’s time and depends on who is doing it.
  • A model asked directly gives round numbers that do not know the team’s rates, risks or way of working.

The team needed estimates that come fast and still follow its own rules.

Solution

The model does the breakdown, and the team’s knowledge shapes it: rates by role and grade, overhead and risk coefficients, base tasks and project templates all go into the prompt. A manager gets a first estimate in a minute and corrects it like a spreadsheet.

Estimate

From a sign-in to a priced breakdown of the project on one page.

Sign-in. The cards on the left show what the service produces: hours, timeline, a CV score.
Sign-in. The cards on the left show what the service produces: hours, timeline, a CV score.
Two jobs to start from: estimate a project or analyse CVs to put a team together.
Two jobs to start from: estimate a project or analyse CVs to put a team together.
The project is described in plain words. Examples under the field show how much detail is enough.
The project is described in plain words. Examples under the field show how much detail is enough.
A screenshot or a document can go along with the text.
A screenshot or a document can go along with the text.
The estimate: tasks by category with hours and cost, a risk buffer in each, and the totals on the right. Export or send it for approval.
The estimate: tasks by category with hours and cost, a risk buffer in each, and the totals on the right. Export or send it for approval.
Tasks are added and edited right in the table; the totals follow.
Tasks are added and edited right in the table; the totals follow.

History and profile

Every estimate is kept and grouped by date, so an old project is one click away.

History of estimates in the sidebar, today, this week and earlier.
History of estimates in the sidebar, today, this week and earlier.
Profile and password in one window over the work.
Profile and password in one window over the work.

Control centre

Where the team teaches the model how it works. Each block can be switched on or off, and what is on goes into the prompt.

Hourly rates by role and grade: middle, senior, lead.
Hourly rates by role and grade: middle, senior, lead.
Coefficients for overhead and risks, with a note on what each one covers.
Coefficients for overhead and risks, with a note on what each one covers.
Statistical coefficients come with a definition written for the model, so it reads them the right way.
Statistical coefficients come with a definition written for the model, so it reads them the right way.
Base tasks by competence, with hours for each grade.
Base tasks by competence, with hours for each grade.
Project templates: an online store, a travel site, a landing page.
Project templates: an online store, a travel site, a landing page.
Inside a template: its components, their category, hours and whether they are required.
Inside a template: its components, their category, hours and whether they are required.
The system prompt and the decomposition prompt are edited right here.
The system prompt and the decomposition prompt are edited right here.
The model and the embeddings, with prices per thousand tokens next to each.
The model and the embeddings, with prices per thousand tokens next to each.