PROFILEPRODUCT MANAGER FOCUSMOBILE × GROWTH DOMAINFINTECH / CRYPTO MAU60K+ JTBD INTERVIEWS60+ STACKSQL · A/B · COHORTS · PYTHON STATUS● AVAILABLE LOCATIONREMOTE / GLOBAL PROFILEPRODUCT MANAGER FOCUSMOBILE × GROWTH DOMAINFINTECH / CRYPTO MAU60K+ JTBD INTERVIEWS60+ STACKSQL · A/B · COHORTS · PYTHON STATUS● AVAILABLE LOCATIONREMOTE / GLOBAL
DMITRII SHABANOV / 002 FILE: PM_PROFILE.v3 UPDATED · 2026

Mobile & Growth Product Manager.

I help fintech and crypto products grow by fixing the hardest part of the funnel — activation and first value. At Veles.finance I cut time-to-first-trade by ~7x and tripled new-user deposits — by turning user behavior into measurable revenue, not feature lists.

REC · SUBJECT_001 · ON FILE
// ROLEPM · Mobile / Growth
// PRODUCTveles.finance
// MAU60,000+
// STATUS● open to roles
§ 01 — About

The pitch.

I treat onboarding and first value as core product mechanics — not marketing add-ons. My focus isn't shipping features. It's shipping levers that move metrics.

I build for the moment a user goes from curious to committed. That's where most fintech products quietly bleed revenue — and where I'm most useful.

I diagnose growth problems through SQL-driven analytics, not assumptions. I redesign flows based on user research, not opinion. I validate with A/B tests, not gut feel. At Veles, that approach took new-user activation from a passive metric to a system — cutting time-to-first-trade by ~7x and lifting bot-launch success from 88% to 96.5%.

I'm especially effective in environments where data is available but underused — where growth is expected, but the system for experimentation is missing.

Activation First-value Monetization Funnels Cohorts SQL A/B testing Onboarding Mobile Crypto Fintech LLM-native
§ 02 — Selected work

How I move metrics.

Case 01
Activation
SHIPPED · 2025

Cutting time-to-first-trade by ~7x.

// Veles.finance · new-user onboarding redesign
// Context

New users were dropping off between sign-up and their first bot launch. Most never made it to first value. The flow was working — but slowly, with high error rates and no clear diagnosis of why.

// What I did

Ran 14 JTBD interviews and 4 UX tests with new users to map where intent breaks. Rebuilt the onboarding flow around the moments users actually got stuck — API connection, configurator defaults, first bot setup. Wired up event tracking in Firebase + GA and funnels in Grafana to measure each step before and after.

// Result

Time from sign-up to first bot launch dropped ~7x. Share of bot launches with config errors fell from 12% to 3.5%. Onboarding now operates as a measurable system — not a black box.

~7x
Faster
time-to-first-trade
12% → 3.5%
Bot-launch
error rate
18
User sessions
(JTBD + UX)
Case 02
Monetization
SHIPPED · 2025

Tripling new-user deposits.

// Veles.finance · bot showcase rebuild
// Context

The bot showcase — the first place new users meet the product — was unfiltered and noisy. Users picked low-performing bots, hit minimum-deposit limits that didn't match strategy requirements, and churned before seeing any trade work in their favor.

// What I did

Curated the showcase down to bots with proven performance windows. Tuned minimum deposits to match each bot's actual capital requirement — based on cohort data, not assumptions. Layered in category tabs and bot-detail context so users could pick by intent, not by luck.

// Result

New-user deposits grew 3x. Registration-to-first-trade conversion lifted by +8.8 p.p. The showcase became the platform's strongest first-value mechanic.

3x
User
deposits
+8.8pp
Reg → trade
conversion
+9.8%
Avg
deposit size
§ 03 — Track record

Numbers that compound.

Active user base
60K+
MAU on Veles.finance — #1 in segment among algorithmic trading platforms.
Onboarding speed
~7x
Faster time-to-first-trade after onboarding redesign — activation now measurable.
New-user deposits
3x
Growth after showcase rebuild — paired with +8.8 p.p. reg → trade conversion.
User research
74+
JTBD interviews and UX sessions across two products — segmentation, churn, activation.
§ 04 — Experience

Where I've shipped.

2024 — Present NOW

Mobile Product Manager — Growth

Veles.finance · algorithmic trading platform #1 in segment · 60K+ MAU

// Ownership

  • Own the mobile product (iOS & Android) end-to-end — from discovery and BRD to release and post-launch iteration
  • Shipped backtests, multi-pair bots, push, in-app chat (Jivo), and UX improvements across the configurator and cabinet
  • Initiated the A/B testing process on mobile — systematic hypothesis validation across the product

// Impact

  • Onboarding redesign: ~7x faster time-to-first-trade, bot-launch errors 12% → 3.5%
  • Showcase rebuild: 3x new-user deposits, +8.8 p.p. reg → trade conversion
  • Built product analytics from scratch — Firebase, GA, Grafana, SQL

// Research

  • 14 JTBD interviews + 4 UX tests with new users — fed directly into onboarding and showcase redesigns
  • Competitor teardown of 10+ platforms (OKX, 3Commas, Bitsgap, Pionex, WunderTrading, Cryptohopper) — 16+ validated hypotheses in backlog

// Ways of working

  • Built lightweight mobile prototypes in Claude Code & Cursor — validated UX ideas without burning engineering hours
  • Set up pre-decomposition for large mobile features — fewer design iterations, shorter cycle from idea to hand-off
Oct 2023 — Nov 2024

Product Manager — Subscription

"Uzhin Doma" · meal kit service Top-3 in market · 70K+ MAU

// Product impact

  • Reduced early churn and increased subscription adoption via web and mobile flow redesigns
  • Owned the product experimentation cycle end-to-end — hypothesis, research, A/B, analysis

// Research depth

  • Conducted 60+ JTBD interviews — refined segmentation and product positioning
  • Translated qualitative findings into testable bets that shaped the roadmap
§ 05 — Method

How I work.

Built with code.
Not just decks.

Most PMs talk about AI. I ship with it daily. I diagnose funnels in SQL, not dashboards. I prototype mobile flows in Claude Code & Cursor before asking engineering to build them. I synthesize user research with LLMs to compress weeks of interviews into days of decisions.

The result: faster diagnostics, sharper hypotheses, fewer wasted engineering hours. When a question takes a single PM 4 hours instead of a team's week, the product moves at a different pace.

SQL · cohorts Firebase · GA Grafana Claude Code Cursor Figma A/B testing JTBD research
~/dmitrii — pm_workflow.py
$ analyze --funnel signup_to_first_trade → Loading cohorts (n=12,481)... → Detected drop-off: api_connection_step ✓ -47% conv. before first bot launch   $ prototype --flow onboarding_v2 --tool claude_code → Generating mobile prototype... ✓ Built in 2h. Validated with 6 users.   $ synthesize --interviews 14 --model claude → Clustering pain points... ✓ 3 themes. 1 actionable bet. ✓ Hypothesis queued for A/B test. $
§ 06 — Foundations

Education & training.

// DEGREE

MSc — Digital Product Management

HSE University, Moscow · 2022 — 2024
// DEGREE

BSc — Economics

HSE University, Moscow · 2018 — 2022
// CONTINUING EDUCATION

Courses & certifications

  • 2024 Ivan Zamesin — How to Build Products
  • 2024 GoPractice — Data-driven product simulation
// 07 — Let's talk

Hire the PM who
moves metrics.

Looking for a Mobile / Growth PM who works hands-on with funnels, cohorts, and experiments? Let's connect.