Looking for a job

Daniil Diveev

@daniildiveev

Message

Product Analyst at Tinkoff who drove a 5% roll-rate reduction through A/B-tested collection communications.

Russian Federation

Message

What I'm looking for

I'm open to data analyst, product analyst, and data science roles, while growing toward ML systems and inference optimization.

At Tinkoff, I designed and ran A/B experiments on collection communications, driving a 5% roll-rate reduction with no cure-rate degradation. At X5 Tech, I ran experiments across app funnels and customer campaigns, and extended a production customer-segmentation model.

I also built Lifemetric’s health-data pipelines and scoring systems, and evaluated RAG-based support chatbots at Yandex. Outside work, I write GPU kernels in Triton and CUDA, including a FlashAttention-style fused attention kernel benchmarked on an RTX 4090.

Experience

Work history, roles, and key accomplishments

Data Analyst

Yandex

Jul 2026 - Sep 2026 (2 months)

- Evaluated the quality of RAG-based LLM support chatbots. Supported ~5 A/B experiments covering new versions of the RAG pipeline, search indexes and prompts, plus a request-topic classifier applied at handoff between bots. Metrics: self-service rate, automation rate, FCR, escalation rate, CSAT.
- Used LLM-as-a-judge scores, cross-checked them against human annotations and flagged discrepancies.

Data Engineer

Lifemetric

Oct 2025 - Jul 2026 (9 months)

LifeMetric is a health platform that turns Apple Health and wearable data (Apple Watch, Whoop, Oura) into five daily lifestyle scores: Move, Eat, Relax, Sleep and Balance.

- Built the core pipeline that aggregates raw health data into multi-level time series (from 5-minute to monthly), with timezone-aware processing and safe re-runs
- Designed the storage schema for raw health samples and select

Product Analyst

Tinkoff

Feb 2026 - Jun 2026 (4 months)

- Designed and ran A/B experiments on multi-channel collection communication scenarios (chat, calls): hypothesis design, MDE and sample-size estimation, statistical testing in Python and Greenplum

- Drove a 5% roll-rate reduction with no cure-rate degradation; winning variants shipped to production

Data Analyst

X5 Tech

Sep 2025 - Feb 2026 (5 months)

- Owned the full campaign cycle for customer clusters (core, churn-risk, budget-focused): segmentation, experiment design, metric selection, results evaluation.

- Extended and retrained a production customer-segmentation model (multiclass classification, gradient boosting) on aggregated behavioral features.

Product Analyst

X5 Tech

Apr 2025 - Sep 2025 (5 months)

- Designed and ran 15 A/B experiments across app funnels; shipped changes drove ~7% conversion uplift in target funnels.

- Defined and instrumented ~50 key product events (onboarding, activation, purchases) enabling reliable product analytics and funnel tracking.

Software Engineer

elDinero

Oct 2024 - Apr 2025 (6 months)

- Built a pipeline that automates review replies using sentiment and topic analysis, cutting manual answering by 90%

Data Analyst

X5 Tech

Jun 2023 - Apr 2025 (1 year 10 months)

- Built and automated 140+ recurring marketing reports (conversion, average check, revenue) in Python, PySpark and Greenplum, cutting preparation time by ~90%.

- Developed 5 data marts and 3 ETL pipelines over transactional and loyalty data for marketing and product teams.

- Introduced a unified campaign design framework: automated MDE pre-estimation and result logging for standard campaigns,