Dmytro

Phone: [open contact info](look above in the "contact info" section)
Dmytro Dorychevskyi
Telegram: @d_dorichevskiy
QA Lead / Automation Linkedin: Dmytro Dorychevskyi
Email: [open contact info](look above in the "contact info" section)
City: Kyiv
About Me

QA Lead with 5+ years in QA, specializing in test automation and AI-assisted quality workflows. I

build and operate a Claude Code-based agentic QA pipeline optimized from 7.17M to 55.6K
T oo ls a d T h n ec no o l g s ie

tokens per run (129×) through tool selection and model-size decisions. AI & Ag n i T ing
e t c ool :

I’m also data-driven about defect prevention: introduced a root cause analysis process for

production and staging escapes — each missed defect is traced back to why testing didn’t catch
C C
laude( ode custom agents,

slash commands, subagent
it, tightening the test suite over time.

orchestration, hooks, memory

Work Experience
management )

LLM -assisted testing work fl
ows

QA Lead – Shelterzoom
Prompt engineering for test
Proficiently contributed to diverse commercial projects and company marketing, showcasing
automation agents
expertise in UI/UX.

Projects:
La ngu g a es & Scr ip ing
t :

Mithra (current) — enterprise AI document intelligence platform (React 19, TypeScript, AWS J Sava cript / T Sype
cript

Bedrock RAG); users query document libraries via natural language with citation-grounded
Python

responses.
Bash / L inux console
Document GPS (earlier) — blockchain-based D M
R platform for email attachments; permission

controls, real-time access tracking, and e-signatures across O utlook, Gmail, i O S, and Android. Test ing T ools & Frameworks:

R p n ibi i i
es o s l t es: Playwright

O wn QA strategy across the full product lifecycle: re q uirements analysis, test design, release Vitest / Testing L ibrary

sign-o ff
, and post-release monitoring
Postman

E A A ;
ATDD
xecute hands-on testing across web, PI, and I response layers maintain and extend

Playwright E2E automation covering key user fl ows

Lead and mentor a 7
-person QA team ; partner with dev on unit test coverage strategy and
Process & Doc u me n tat o i n :

R

ownership boundaries

A uthor technical investigations on testing strategy, tooling candidates, and A I agent
T est ail

architecture Jira

K y
e Ach e i v eme nts: G t
i

AI Cost Optimization
G L C

it ab I

129× token reduction (7.17M → 55.6K tokens/run) — diagnosed MCP tool overuse; enforced C D T
hrome ev ools

playwright-cli bash commands

70–80% further reduction by switching Opus → Haiku (5× cheaper/token)

O A S
pen PI / wagger

A/B experiment (JSON Model vs Pure Snapshots) with data-backed scalability crossover
recommendations E d a uc tion
Agentic QA System

Built Claude Code pipeline: orchestrator, AI bug reporter, Jira integration, test case generator, Ps y cholo gy / B achelor

Playwright runner — full autonomous QA coverage
National Pedagogical D ragomanov
QA Process

C ross-platform DRM validation (O utlook, G mail, i OS A , ndroid ) for enterprise clients in financial,
University

legal, and real estate

R oot cause analysis per production/staging escape — each missed defect traced to a

coverage gap, suite tightened iteratively

L a g ag s
n u e

C ombined automation + exploratory regression process — shorter release cycle, maintained English — C2 Proficient Certificate

coverage depth

QA Engin eer – G niu
e see

Projects:

S ocial network for elderly people (W eb / M obile )

C hrome extension for crypto wallet

M obile app for measuring sun intensity

W eb application for domestic airlines

R p n ibi i i
es o s l t es:

O wned test documentation: strategy, test plans, cases, checklists, and bug reports

E xecuted regression, integration, and E2E test cycles

T ested third-party integrations: S tripe, Plaid, customer.io, Pusher

D elivered cross-platform coverage across web and mobile ( OS A
i / ndroid )

C ollaborated with dev team on re q uirements analysis, troubleshooting, and root cause

investigations