We check that the product works, translate it for a new audience and keep quality up after release. A Tashkent studio with clients in Russia, Hong Kong, Cyprus and Armenia.
Tell us about your taskTeams that already have a product but no dedicated QA — or a QA team that cannot keep up with releases and new markets.
Web and mobile apps before release, after major changes and when entering a new country.
Betting and gaming platforms, high-load services with frequent promotions: regression before launches, load tests, promo material checks against platform rules and regulator requirements.
Analytics and performance marketing services: integration testing, automated reporting and lead handling.
Catalogs with tens of thousands of SKUs, bots for customer requests, data freshness control.
Four practices, one goal: the product works as designed for every user in every market.
Functional, regression, load, cross-browser and acceptance testing of web and mobile products. We verify the product against requirements and user expectations before release and after every major update.
We work in your tracker or ours: every defect comes with steps to reproduce, environment, severity and priority, a screenshot or screen recording. Blockers are escalated the day they are found.
Translation of interfaces, documentation, emails and marketing materials into Russian, English and Uzbek; other languages with vetted partners. We maintain a glossary and translation memory so terms stay consistent across releases.
LQA — linguistic testing: we check the translated product in the interface, not in a spreadsheet: clipped strings, wrong inflections, dates, currencies, formats.
First-line user support under the client's playbook: chats, email and messengers, escalation to the dev team, knowledge base.
Content moderation and pre-publication checks: compliance with platform rules, advertising requirements and internal standards.
How we do the first three faster. We build data validation pipelines, bots for collecting requests and reports, integrations with trackers, CRMs and spreadsheets, and AI assistants on modern language models.
Separately — automating routine in your team: lead handling, stock reconciliation, document preparation.
A fixed deliverable at every step — from the first call to the signed act.
Describe the task in the form or by email. A 20–30 minute call to clarify product, markets and timing.
We describe the scope, acceptance criteria and report format, then send an estimate and schedule. The estimate is free.
Weekly reports and tracker access: coverage, defects with severity and priority, translation progress — in real time. Blockers get a separate response SLA.
Final report, act, hand-over of artifacts. By agreement we stay on for release support.
Some clients are not disclosed under contract terms — we name the industry and the result.
Spend-to-top-up reconciliation across every account to the cent, around 200 automated checks of internal tools in pytest, scheduled jobs with chat alerts and an MMP access audit. Python, FastAPI, SQLite, ad network and MMP APIs.
Outcome: a loss of part of the spend in the panel and double-counted registrations in the export found and fixed, all working sheets corrected in one run, about half of the excess accesses closed.
Read the case →We built a full pre-promotion check cycle: regression of key flows from sign-up to withdrawal, runs on iOS, Android and all major browsers, LQA of promo materials directly in the interface in several languages, and a compliance check of every creative against platform and regulator requirements.
Outcome: a prioritized defect report a day before launch — and not a single critical incident on the launches we supported.
Read the case →Marketplace synced with the ERP by warehouse, XML feeds for autoload, an LLM assistant for client chats, bank reconciliation via the bank's Open API. Java 21 and Spring Boot, Node.js and TypeScript, Python and FastAPI.
We prepared the platform for a peak campaign: a load profile from real logs, target thresholds for p95 response time and error rate, peak scenarios several times above normal traffic run until the thresholds broke, API and ad-account integration tests, database and queue monitoring during runs, and a report with reproducible scenarios for the dev team.
Outcome: every bottleneck found and ranked by cost of fixing before launch; the scenario set stays with the team for a re-run after each release.
Read the case →An adapter for every chain's format and a recognition pipeline: the PDF text layer, Tesseract OCR with automatic rotation, Claude Vision for disputed cases. Line items are reconciled with the document total, and every load is reversible. Python, FastAPI, SQLAlchemy, pandas.
Outcome: manual data collection replaced by automatic loading, documents needing manual entry cut fourfold; 376 automated and 12 browser tests.
Read the case →VC LAB is a quality assurance studio in Tashkent. We help teams ship products that work for every user: testing before release, localizing for new markets and keeping quality up after launch.
We work remotely with clients in Russia, Hong Kong, Cyprus and Armenia, in Russian, English and Uzbek. Every project is delivered with documents — test plan, reports and act — so the result is easy to verify and hand over.
Legal entity: VC LAB LLC (MChJ), Tashkent, Republic of Uzbekistan.
We reply within one business day and send an estimate in 1–2 days. Free of charge.