Skip to content
FI
RU EN
Telegram
open to QA/AQA leadership roles Remote only · Moscow, UTC+3

Farid Ismailov QA/AQA Lead

I build QA/AQA around product risk: set priorities, develop engineers, and coordinate work with development and analytics. I combine leadership with test automation architecture and hands-on tooling for the team.

MEDIA-100 LLC · QA Automation Lead

Experience & case studies Get in touch Download CV

A few key facts.

12+ years
in QA

Exactpro · Wiley · Sber

5+ years
accountable for product QA at Sber

B2B/B2C marketplace · development, analytics, product

Android · iOS
mobile automation

Kotlin · Appium · BrowserStack

experience and leadership

Experience.

12+ years in QA, including more than five years accountable for quality at a B2B/B2C marketplace within Sber. I now lead mobile automation at MEDIA-100 LLC and stay hands-on.

  1. Jun 2026 — present · ongoing

    QA Automation Lead — Mobile · MEDIA-100 LLC

    I lead Android and iOS test automation: set direction, organize engineering work, and build the team's tools. With manual QA and development, I define which scenarios come next, what automation needs, and who resolves the blockers.

    • Developed the QA/AQA strategy: what to automate first, how to assess regression quality, and when to scale capacity, based on a review of coverage, the backlog, and business risks
    • Defined the manual QA-to-AQA handoff standard: priority, precise steps, data, and an unambiguous expected result. Set out which tasks the team can start and what needs preparation first
    • Organized engineering work through priorities, review, and daily updates. Defined failure triage that separates product defects from test, data, and infrastructure issues, assigns ownership, and verifies fixes
    • Built two AI tools for the team: a work-chat QA assistant and an agent-assisted test development environment. Delivered guidance, ran demos, and trained two engineers on real tasks
    highlights · stack · MEDIA-100 LLC
    • Defined and agreed on a Jira workflow for locator requests: AQA describes the problem, development plans the fix, and AQA verifies it. Separately specified requirements for UI identifiers, test data, and mocks
    • Set review rules that keep task-related fixes in the current MR and move existing technical debt to the backlog, giving reviews a clear acceptance boundary
    • Authored technical requirements for AQA lead and senior engineer job descriptions; HR incorporated the changes, covering skills, quality expectations, and team responsibilities

    Kotlin · Appium · BrowserStack · TestIT · Android · iOS · CI/CD · Allure · Jira · MCP · Codex · AI Agents

  2. Sep 2020 — Apr 2026 · 5 yrs 8 mo

    QA Lead / Lead SDET · Sber

    B2B/B2C marketplace within Sber. Accountable for QA direction and release readiness in a team of 17+ developers. Combined technical leadership and engineer development with close coordination across development, analytics, and product.

    • Was a key source of product expertise: onboarded new hires across development, analytics, and QA, explaining business logic, integrations, and system behavior. Helped the team work through requirements and complex product scenarios
    • Participated in hiring and evaluating QA engineers, developed their testing and automation skills, and reviewed test code
    • Built QA processes and automation from scratch: created a Python framework, migrated it to Kotlin to match the development stack, and integrated runs into CI/CD
    highlights · stack · Sber
    • Established QA infrastructure: regression suites, UAT/PSI, release-readiness criteria, and quality gates
    • Covered smoke, regression, and end-to-end flows for product setup, sales, statuses, reports, and integrations
    • Tested backend and APIs across REST, JSON Schema, business rules, and SQL validation; diagnosed defects through logs and traces
    • Ran API and integration performance checks with JMeter and Yandex.Tank, tracking latency, throughput, error rates, and endpoint regressions

    Python · Kotlin · Java · Selenium · Playwright · Appium · REST API · Postman · RestAssured · SQL · PostgreSQL · MySQL · GitLab CI · Jenkins · Docker · OpenShift · Kubernetes · JMeter · Yandex.Tank · Allure · Jira · Confluence

  3. Nov 2016 — Aug 2020 · 3 yrs 10 mo

    Senior QA Engineer · Wiley

    International enterprise: online learning platforms, scientific journals, and educational content distribution systems. Distributed team, English-speaking colleagues.

    • Built test infrastructure: test cases, test plans, requirements traceability matrices, regression coverage
    • Developed and maintained UI automation in Selenium WebDriver + Java / Python
    • Tested REST APIs and integrations between platform components
    highlights · stack · Wiley
    • Verified data and diagnosed defects across Oracle, MySQL, PostgreSQL, MS SQL
    • Worked in CI/CD: Jenkins / GitLab CI, automation integration, failure analysis, defect prioritization
    • Test leadership: requirement decomposition, effort estimation, coordination of testing activities

    Java · Python · Selenium WebDriver · TestNG · JUnit · Allure · REST API · RestAssured · Postman · SQL · Oracle · MySQL · PostgreSQL · MS SQL · Jenkins · GitLab CI · Jira · Confluence

  4. Jul 2014 — Nov 2016 · 2 yrs 5 mo

    QA Engineer · Exactpro Systems

    Financial systems: post-trade infrastructure, clearing, settlement, and two-factor authentication. Worked on critical exchange and clearing-house services where calculation accuracy and continuity matter.

    • Functional, integration, regression, smoke and exploratory testing
    • Designed test scenarios from project documentation and financial-domain requirements
    • Tested 2FA scenarios using mobile applications
    highlights · stack · Exactpro Systems
    • Validated UI and data through SQL
    • Worked in Jira / Redmine, TestRail, Git / SVN
    • Tested 24/7 critical post-trade systems for an exchange and a clearing house, with strict requirements for calculation accuracy and service continuity

    Java · Selenium · Selenide · TestNG · Allure · SQL · Oracle · MySQL · Postman · HTML · CSS · JS · Git · SVN · Jira · Redmine · TestRail

featured case studies

Systems I've built.

These cases show both sides of my work: tools and processes built for teams, and products I build on my own. Each has an overview and technical walkthrough; work systems are anonymized.

Work project

Mobile QA/AQA Development

Developing mobile AQA through risk-based priorities, team workflows, and device verification.

Lead automation and build the tooling

After reviewing coverage, the backlog, and business risks, I developed the mobile AQA strategy: automation priorities, regression assessment, and scaling criteria. I organize engineering work, set case-readiness requirements, and work with manual QA and development to identify what needs preparation.

What the system does

  • Delivered an agent-assisted test development environment and team guidance.
  • Agreed with development on a Jira workflow for locator requests; separately specified UI identifier, test data, and mock requirements.
  • Trained two engineers; runs are checked on BrowserStack devices with app and OS details.
  • Authored technical requirements for AQA lead and senior engineer roles; HR accepted the changes.
Architecture and technical details · Mobile QA/AQA Development
  • Developed the QA/AQA strategy: automation priorities, regression assessment, and scaling criteria based on coverage, backlog, and business risks.
  • Set the manual QA-to-AQA handoff standard: priority, precise steps, data, and expected result; defined which cases were ready and what needed preparation.
  • Delivered Codex guidance and skills for the workflow from case analysis and code reuse through implementation, review, and device verification.
  • Agreed with development on a Jira workflow for locator requests; separately specified UI identifier, test data, and mock requirements.
  • Connected issue, documentation, test-case, and code sources through MCP; use BrowserStack runs and TestIT results for verification.
  • Trained two engineers on real tasks; organize review and failure triage with assigned ownership and fix verification.
Mobile AQA · from case to device verification Click to expand
Mobile AQA · from case to device verification

Scroll horizontally to explore the diagram

Kotlin · Appium · BrowserStack · TestIT · Allure · GitLab CI · Codex · MCP

An anonymized work project under NDA.

Work project

E2E Automation · Web and PWA

Web-admin and PWA automation with failure analysis and Jira bug drafts.

Framework and AI-tool author

This work case covers an internal web admin and mobile PWA across several environments, where I owned automation and release readiness. I connected failed-run logs from S3 MCP with project documentation in CRAG; the AI assistant prepares a Jira draft that I review and complete. The team can also search the documentation from the IDE.

What the system does

  • Built a Kotlin E2E framework and integrated regression, TestIT/Allure reporting, and release-readiness checks with CI/CD.
  • Cross-checking run logs with documentation prepares a Jira draft for human review.
  • Made project-documentation search available to the team from the IDE.
Architecture and technical details · E2E Automation · Web and PWA
  • Kotlin 1.9 + JUnit 5 + Playwright for Java + Allure
  • End-to-end coverage for an internal admin web panel and a mobile PWA
  • Runs in staging / pre-prod / prod with restricted-egress build agents
  • Local Playwright browser mirror handling for offline build hosts
  • TestIT / Allure reporting, regression and release-readiness gates
  • CRAG as the project's documentation index — specs, ADRs, runbooks, domain guides
  • AI triage: agent fetches logs via S3 MCP, cross-checks with CRAG, drafts a Jira bug
  • Project-docs search from the IDE: "how does X work?" returns relevant doc snippets
Architecture · E2E runs and AI-assisted failure analysis Click to expand
Architecture · E2E runs and AI-assisted failure analysis

Scroll horizontally to explore the diagram

Kotlin 1.9 · JUnit 5 · Playwright for Java · Allure · TestIT · Gradle · GitLab CI · CRAG · S3 MCP · Jira API

A work project under NDA; the public name is generalized and product details remain confidential.

Personal project · Flagship personal project

Self-hosted AI Platform

My flagship personal AI platform: four companions with memory, voice, and tools.

I design, build, and run it myself

What the system does

  • Combines conversation, research, documents, images, audio, and video.
  • Keeps searchable, versioned long-term memory linked to source material.
  • Runs tasks through specialist agents, produces voice replies, and exposes execution in an operator console.
Architecture and technical details · Self-hosted AI Platform

It started as a Telegram chat bot and grew into a platform for four bots with different roles. They remember conversations, find information, work with documents and recordings, use tools, and reply by voice. I also built a web console for the archive and for inspecting what an agent did.

  • Shared runtime for four bots: model routing, tool-call loops, specialist roles, and continuation after selected provider failures.
  • Durable external-operation receipts distinguish successful, partial, failed, and unknown outcomes; ambiguous calls are not blindly replayed.
  • Multimodal pipeline sends fitting audio/video as native input and processes larger recordings in chunks, retaining coverage and partial results.
  • SQL-first memory uses SQLite + FTS5, attribution, versions, and addressable sources; auxiliary indexes remain projections.
  • New memory is available immediately; extraction shares the summary pass and semantic maintenance runs as a bounded nightly job.
  • Work-context search brings connected events, documents, decisions, and unresolved questions into an operator interface.
  • Voice replies, grounded web research, and code execution are part of the runtime; connected tools use MCP and adapters, while code validates their parameters and allowed actions.
Self-hosted AI Platform · architecture Click to expand
Self-hosted AI Platform · architecture

Scroll horizontally to explore the diagram

Bots sharing the runtime
4

Python · Telegram API · OpenRouter · Gemini · SQLite / FTS5 · MCP · Multimodal · TTS · Docker · Linux / systemd

Actively developed personal product. Architecture and an anonymized demo are available in an interview.

Personal project

Not So Super Whisper

Speak instead of typing, straight into the app you need.

Build the macOS and Windows versions

What the system does

  • Shows live transcription and buffers opening speech while the model loads.
  • Inserts text into the focused app or saves it as a note.
  • The Windows Island combines dictation, recent notes, and system notifications.
Architecture and technical details · Not So Super Whisper

I built hotkey dictation: speak, watch text appear as it is recognized, and insert it into the focused field. Alternatively, save it as a Markdown voice note. On Windows, I added a compact Island for dictation, notes, and notifications.

  • Streaming Nemotron uses FluidAudio / CoreML on macOS and Sherpa-ONNX / CUDA on Windows; Silero VAD filters silence.
  • Capture buffering preserves ordered audio until the model is ready.
  • Global hotkey, quick insertion, and clipboard / Markdown output without opening an editor.
  • Local ASR is separate from optional cloud text polishing; raw transcription survives polish failures.
  • A separate Gemini Direct route uses cloud speech recognition.
  • Windows Island provides dictation, notes, and a notification inbox, with restorable banner settings.
Architecture · dictation flow + platform-specific runtimes Click to expand
Architecture · dictation flow + platform-specific runtimes

Scroll horizontally to explore the diagram

Swift · AppKit · CoreML · C# / .NET · WPF · Win32 API · Nemotron · Sherpa-ONNX · CUDA · OpenRouter

Native Windows and macOS implementations. Windows ASR requires NVIDIA CUDA; the Island is Windows-only.

Work project · Operation paused

Team QA Assistant

A work-chat assistant for meeting summaries and analysis of requirements, MRs, and CI failures.

Built the agent and connected work sources

What the system does

  • Built and used in the team's work chat; operation is currently paused.
  • Prepared meeting summaries and analyzed requirements, MRs, and CI using connected sources.
  • Used read-only work-system access and searchable discussion history.
Architecture and technical details · Team QA Assistant

I built and used a QA agent in the team's work chat; operation is currently paused. It helped analyze meeting recordings, requirements, code changes, and CI failures by reading connected issue, test, documentation, and Git systems and retaining discussion history.

  • Supports requirements, test-case, code-change, and run-result analysis.
  • Adapters connect Jira, TestIT, Notion, Git, and chat within the granted access.
  • The agent reads connected systems but cannot change them.
  • History and summaries live in SQLite with full-text search.
  • Incoming events are verified with HMAC and deduplicated. Context is separated by author, chat, and thread.
  • Reviews show the material used; missing code or sources are reported explicitly.
Architecture · Work-chat QA Assistant Click to expand
Architecture · Work-chat QA Assistant

Scroll horizontally to explore the diagram

Python · FastAPI · Gemini · SQLite / FTS5 · Webhooks · MCP-style adapters

Used in the team's work chat; operation is currently paused.

Personal project

Conference Capture

After the meeting: a recording, transcript, and what was on screen.

Build the extension and local service

What the system does

  • Records tab audio, a separate microphone track, and optional video.
  • Combines the post-meeting transcript with screen-sharing context.
  • Exports media, meeting summaries, and notes to a chosen folder and Markdown.
Architecture and technical details · Conference Capture

I built a Chromium extension and a local Python service. They record browser meetings, then process speech and screen content separately. The output is a timestamped transcript, meeting summary, and notes about the material shown.

  • Separate tab and microphone tracks preserve speech provenance without inventing remote speaker identities.
  • Recording-only by default; live ASR is optional. Transcription and visual analysis use configured cloud providers.
  • Speech timestamps and visual events are fused after independent audio/video processing.
  • Media persists during capture; processing reuses chunk receipts and can resume after a backend restart.
  • Partial processing retains completed transcripts and explicit status; closing the browser ends further capture.
  • Human-readable export combines summary, transcript, and screen notes in one meeting.md.
Architecture · post-meeting workflow Click to expand
Architecture · post-meeting workflow

Scroll horizontally to explore the diagram

TypeScript · Chromium extension · WebAudio · Python · FastAPI · WebSocket · OpenRouter · FFmpeg

Personal project. Recording stays local; AI processing uses configured external models.

Other systems and experience

Personal project

AI Judge · Verify-Ship Gate

Checking agent changes and finding out why answers got worse.

Verification-system author

What the system does

  • Binds verification to exact source bytes and a change set.
  • Checks affected contracts, callers, and adjacent useful behavior.
  • Supports replayable answer evaluations and regression analysis.
Architecture and technical details · AI Judge · Verify-Ship Gate

I built code checks and answer evaluations for my AI platform. Tests check program behavior; separate examples help compare the meaning and quality of answers. Source material and results are retained so a check can be repeated and a regression investigated.

  • Source identity and recorded outcomes prevent an old run from standing in for a new change.
  • Fixtures cover routing, tools, context handoff, and partial-result preservation.
  • Replayable evaluation cases and review evidence cover semantic answer quality.
  • Change checks consider direct consumers, configuration, and runtime contracts, proportionate to reachable risk.
Change verification · architecture Click to expand
Change verification · architecture

Scroll horizontally to explore the diagram

Python · pytest · JSON Schema · SHA-256 · LLM evaluation

A subsystem of the personal AI platform. Verification design and regression examples are available in an interview.

Personal project

Hassio Walkie

A mobile voice interface to the personal AI platform.

Author, owner

What the system does

  • Push-to-talk from a Kotlin app.
  • Streaming voice replies with text highlighting.
  • A voice channel to the agent and Home Assistant.
Architecture and technical details · Hassio Walkie

I built a mobile voice client for my AI platform: press a button, speak, and listen to the reply. Text is highlighted in sync with the speech. The same channel connects Home Assistant for voice control of devices.

  • Kotlin + Jetpack Compose client with a push-to-talk interface.
  • Streaming TTS and speech-synchronized text highlighting.
  • Voice requests to a personal backend and Home Assistant integration.
Architecture · voice loop and karaoke rendering Click to expand
Architecture · voice loop and karaoke rendering

Scroll horizontally to explore the diagram

Kotlin · Jetpack Compose · Material 3 · Gemini Live · Streaming TTS · WebSocket · Home Assistant API · OkHttp

The mobile client and integrations form a separate subsystem of the personal AI platform.

operating principles

How I work.

A few working habits I rely on day to day.

  • Quality has a business purpose.

    I start with what must not break for users and the business. That drives team priorities, test depth, and release decisions.

  • Working means verified.

    Generated code is not proof of an outcome. Run it through its real consumer and keep the execution evidence.

  • Every task needs an owner.

    I agree on inputs, responsibilities, and acceptance with partner teams. Engineers need to know what comes next and which decision they are waiting for.

  • Build a stronger team.

    I explain decisions, work through difficult problems with engineers, and share the tools. A good process gives people the confidence to work independently.

stack

Stack.

The main technologies I have used in teams and personal projects.

Languages, APIs & data

Core engineering tools for automation and integrations.

  • Python
  • Kotlin
  • Java
  • REST / WebSocket
  • SQL
  • PostgreSQL
  • MySQL
  • Oracle / MS SQL

Automation & quality

Mobile, web, and API checks, reporting, and performance.

  • Appium
  • BrowserStack
  • Playwright
  • Selenium
  • JUnit / TestNG / Pytest
  • RestAssured
  • Allure / TestIT
  • JMeter / Yandex.Tank

CI/CD & infrastructure

Test pipelines, environments, and service operations.

  • GitLab CI
  • Jenkins
  • Docker / Compose
  • OpenShift
  • Kubernetes
  • Linux
  • Nginx / systemd

AI & agent systems

Team tooling and my personal R&D platform.

  • MCP
  • Codex
  • Gemini / OpenRouter
  • RAG / CRAG
  • SQLite / FTS5
  • ChromaDB / Neo4j
  • ASR / TTS
  • Multimodal AI

Native applications

Personal products for dictation and everyday work.

  • Swift
  • macOS
  • C# / .NET
  • Windows
  • NVIDIA CUDA
  • Nemotron / Silero VAD

Workflow & documentation

Tasks, requirements, test cases, and product knowledge.

  • Jira
  • Confluence
  • TestIT
  • TestRail
  • Git
  • Code review
Other tools I have worked with

Postman · JSON Schema · pytest-xdist · Bash / PowerShell · VPN · WaveSpeed · Telethon · Function calling · LLM routing · Context engineering · Agent skills / prompts · Cursor · VS Code · IntelliJ IDEA · Redmine · GitHub

contact

Get in touch.

Open to QA/AQA leadership roles focused on developing the team and processes, automation, and release quality. I stay hands-on as part of the role.

GitHub LinkedIn Discord: leviann