Valeriia Kaimashnikova

AUTODOC grew sales revenue from €1.13 billion to €1.81 billion between 2022 and 2025 (publicly reported), serving 18.9M annual orders and 9.3M active customers.

What I owned

• Product search from retrieval to ranking — Elasticsearch relevance tuning, query understanding, autocomplete, OE/part-number search, and promotional listings.
• The search analytics — built dedicated data marts and defined metric and attribution logic myself, including incrementality (measuring what search truly added, net of cannibalization), with anomaly alerts, YoY comparison, and a Metabase self-service tool.
• A feedback engine spanning search logs, user testing, customer support channel, and AI-summarized Trustpilot reviews — so every backlog item was backed by data, not opinion.
• Close collaboration with Data Science, Engineering, Analytics, and Category/Assortment, Marketplace teams.

Achievements

• Grew conversion and YoY revenue by evolving search from rule-based sorting to ML ranking (CatBoost), with continuous search relevance work. Add-to-cart conversion rose 8% on web and 10% in the app year on year.
• Cut the zero-results rate by 16%. Analysed search logs and built my own classification of failing queries, then worked through the causes: vocabulary gaps where customers searched by identifiers the engine did not handle, such as EAN, misspellings and missing synonyms. EAN (aka GTIN) search was an insight that came purely from data and brought the company €1M in additional annual revenue.
• Ranked 1P, third-party marketplace (3P), and private-label offers in a single search results — balancing buyer relevance with marketplace health (margin, seller offers, own brands).
• Cut time-to-market using AI prototyping — validating clickable, stakeholder-tested prototypes before engineering, shipping with minimal rework.
• Built an LLM-as-a-judge to evaluate search relevance and ranking quality.
• Delivered a promotional-listing tool giving promo managers one smart cross-platform link, backed by product retrieval, sorting, and car-compatibility logic.

• Ran B2B lead generation for a seven-brand group. One three-month campaign delivered 172 SQLs at $4.30 per lead and fed a tender win on a 72-container export deal.
• Built a mobile app, an e-commerce site and five corporate sites on no-code tools. This work led into product.

• B2B martech SaaS for call tracking and paid-channel attribution, the category next door to today's sales intelligence platforms.
• Ran industry rankings across three countries in 2019, with 570 participating agencies and 110 marketing experts on the jury.