Pitch me
Boosted user acquisition and core metrics via high-impact Product Hunt releases and UX improvement
Worked at Pitchme in the product team, collaborating with PMs, developers, and data scientists to design and launch digital products. Contributed to UX/UI design and aligned solutions with user needs and business goals
- Duration2020 — 2023
- RoleLead Product designer
- Work & Impact
- Improved design workflows
- Refined brand identity and visual style
- Tested product hypotheses
- Designed UI mockups
- Designed UI animations
- Built user flows
- Analyzed data for bottlenecks
- Researched user sessions
- Grew key metrics
- Optimized sales funnels
- Enhanced UX copy and microcopy
- Created UI component library
- Key Highlights
- $1.5 million in funding
- Target Markets: UK, EU, US
- #5 Product of the Day on Product Hunt
- International team
- 200% New user growth
- #5Product of the Day on Product Hunt
Selected clients
Project in media
When I joined the company, all legacy mockups were created in Sketch, with Zeplin used for developer handoffs. Design files were stored locally, and tasks were managed in Jira.
Files organization
After aligning with developers on Zeplin’s limitations, I spearheaded a full transition to Figma, leveraging its native Dev Mode for seamless handoffs. To streamline workflows:
Standardized template:
Impact: Streamlined workflows and ensured cross-functional transparency
Standardized template:
- Jira-Figma integration: Linked tickets directly to design files via URLs, tagging files with task IDs
- File cover: Displayed task name and tags (priority, status e.t.c)
- Dedicated pages: "Work in Progress", "Exploration", "Handoff" (developer-ready specs), and "Copy" (for copywriter collaboration)
Impact: Streamlined workflows and ensured cross-functional transparency
Workflow
We also agreed on a unified design process structure
- All projects started from a centralized Figma template
- Final screens merged into master files or replaced legacy versions
- Deprecated assets systematically archived
Key benefits
- For PMs: Instant access to task-specific files with clear version history
- For Team: Guaranteed source-of-truth accuracy with no outdated duplicates
- For Onboarding: 50% faster ramp-up for new members (structured templates + archived context)
Uncovered critical UX gaps via market analysis, session replays, and CustDev, identifying onboarding as the key conversion bottleneck
- Manual Dependency: Sales team intervened to complete registrations
- Low Completion Rates: <30% of users reached "Aha!" moments
- Downstream Impact: Poor data quality hurt matching efficiency
Job story
Employer journey
"When I need to hire fast, I want to post requirements without friction, so I can get quality candidates sooner."
Candidate journey
"When I’m job hunting, I want to showcase my skills quickly, so recruiters notice me."
"When I need to hire fast, I want to post requirements without friction, so I can get quality candidates sooner."
- Redesigned vacancy creation as progressive Q&A
- Auto-extracted skills from job descriptions
Candidate journey
"When I’m job hunting, I want to showcase my skills quickly, so recruiters notice me."
- Profile builder with smart suggestions from resumes/LinkedIn
Figma file
Candidate flow
The candidate onboarding flow suffered from critical UX issues:
1. Poor Transparency
2. Missing Value Proposition
3. Algorithmic Data Gaps
Social media linking (for cybervetting) was:
Identified 3 onboarding failure points:
✓ Cognitive overload from multi-action screens
✓ Absent value communication at decision moments
✓ Critical algorithm inputs (social links) treated as optional
1. Poor Transparency
- Unclear step-by-step progression and action objectives
- Overloaded single-screen design requiring excessive actions
2. Missing Value Proposition
- No visible incentives for profile completion
- Weak communication of benefits (e.g., "Complete 3 more steps to 2x profile visibility")
3. Algorithmic Data Gaps
Social media linking (for cybervetting) was:
- Buried in the flow
- Rarely completed (<15% adoption)
- Lacking explanatory copy
Identified 3 onboarding failure points:
✓ Cognitive overload from multi-action screens
✓ Absent value communication at decision moments
✓ Critical algorithm inputs (social links) treated as optional
Figma file
Employer flow
The employer onboarding flow suffered from parallel UX and technical failures:
1. Core UX Issues
2. Technical Debt
- Forced manual corrections → 2x redundant work
- No match justification
Systemic onboarding flaws:
✓ Non-compliant with HRIS standards (no Workday/Greenhouse parallels)
✓ Document parser achieved only 58% field accuracy
✓ Zero post-completion guidance on next steps
1. Core UX Issues
- No Value Visibility: Failed to communicate "What's in it for me?" during vacancy creation
- HR Mismatch: Flow structure contradicted standard recruiter workflows (e.g., no ATS-style bulk actions)
- Interface Overload: Multi-action screens increased cognitive load
2. Technical Debt
- Document Parsing: File upload feature (intended to speed up onboarding) frequently:
- Forced manual corrections → 2x redundant work
- Final Stage Breakdown: Even compliant users received:
- No match justification
Systemic onboarding flaws:
✓ Non-compliant with HRIS standards (no Workday/Greenhouse parallels)
✓ Document parser achieved only 58% field accuracy
✓ Zero post-completion guidance on next steps
Priority #1 Transforming onboarding to unlock growth
Priority #2 Landing page UX revamp: clarifying core мalue
Priority #3 Enhancing candidate match relevance
Priority #2 Landing page UX revamp: clarifying core мalue
Priority #3 Enhancing candidate match relevance
Employer CJM
Find the best-matching IT candidates based on skill mapping and unbiased scoring
Candidate CJM
Get matched with relevant IT job opportunities based on skills, not demographics
Led end-to-end onboarding redesign with dual flows:
- Employer Path: Streamlined vacancy creation to capture rich role requirements
- Candidate Path: Guided profile completion for higher match accuracy
- System Core: Automated skill mapping from unstructured text inputs
Preparation
Pre-restructuring phase included tracing data dependencies throughout all onboarding touchpoints
1. Data-Matching Dependencies
1. Data-Matching Dependencies
- Which parsed/input fields directly fed the algorithm
- Weight distribution across manual vs. automated inputs
- Top 5 most abandoned form fields
- 37% of employer-required inputs were later unused in matching
First step
As a first experiment, we enhanced the existing employer onboarding with explanatory tooltips to test their impact. While this didn’t significantly improve overall conversion rates, user session recordings revealed:
Key Insight:
Microcopy and guidance alone can streamline behavior—but deeper structural changes are needed to move core metrics.
- Faster progression through data-entry stages
- Reduced time spent per field (indicating improved clarity)
- Fewer misinputs (validated via error-rate reduction)
Key Insight:
Microcopy and guidance alone can streamline behavior—but deeper structural changes are needed to move core metrics.
Candidate onboarding v1
First redesign pass (v1) focused on incremental improvements - retaining core UX while visualizing impacts
Initial Approach
Outcome
Poor performance due to:
Key Insight:
Showing real-time results isn’t enough—progressive disclosure beats density.
Initial Approach
- Minimal screen count with high data density per screen
- Core concept: Instant profile previews after each action
Outcome
Poor performance due to:
- Unresolved cognitive overload
- Perceived complexity (despite improved explanatory copy)
- Low completion rates (<5%)
Key Insight:
Showing real-time results isn’t enough—progressive disclosure beats density.
→
→
→
Step 1
%
15
Step 2
%
9
Last step
%
4
Signup page
%
45
Candidate onboarding v2
Iteration v2 - Stepped Automation
Problem
V1's cognitive overload limited conversions
Solution
Outcome
First measurable lift - 23% completion rate
Key Insight:
Simplified steps drive better conversion, but require:
Problem
V1's cognitive overload limited conversions
Solution
- CV-first data collection
- Multi-step validation
- Hybrid design (transitional tech constraints)
Outcome
First measurable lift - 23% completion rate
Key Insight:
Simplified steps drive better conversion, but require:
- Progress reinforcement (show interim value)
- Further cognitive load reduction
→
Signup page
%
79
%
41
Education
Experience
%
49
CV upload
%
64
Links
%
27
Work expectations
%
22
Last step
%
15
Skills
%
31
→
→
→
→
→
→
Candidate onboarding v3
Iteration v2 - Stepped Automation
Problem
V1's cognitive overload limited conversions
Solution
Outcome
First measurable lift - 23% completion rate
Key Insight:
Simplified steps drive better conversion, but require:
Problem
V1's cognitive overload limited conversions
Solution
- CV-first data collection
- Multi-step validation
- Hybrid design (transitional tech constraints)
Outcome
First measurable lift - 23% completion rate
Key Insight:
Simplified steps drive better conversion, but require:
- Progress reinforcement (show interim value)
- Further cognitive load reduction