Bench+
← Back to talent
K

BENCH+ PROFILE

Ksenia

AI Engineer
Production LLM/RAG services, agentic workflows, AI-backed diagnostics, and backend/data integration

Senior AI Engineer with hands-on production delivery across LLM/RAG services, LangChain-based agentic workflows, Python/FastAPI backends, data integration, validation gates, observability, and marketing analytics automation.

BEST-FIT USE CASE

Building production AI diagnostic or automation systems that combine LLM APIs, retrieval, structured outputs, data pipelines, backend services, release controls, and stakeholder-facing operational workflows.

Not the primary fit: Less aligned with roles focused mainly on model research, mobile development, pure frontend delivery, or infrastructure-only platform ownership.

EVIDENCE HIGHLIGHTS

What supports the assessment

01

Shipped a production LLM/RAG service using Python, FastAPI, LangChain, PostgreSQL with vector search, ClickHouse, Docker, and LLM APIs for marketing and product diagnostics.

02

Designed and shipped single-agent and multi-agent workflows with orchestration, context management, tool calling, retry/fallback handling, and safeguards for retrieved and external API inputs.

03

Created an evaluation set and release-blocking checks for factuality, citations, structured outputs, and unsupported-answer regressions.

04

Built AI reporting and measurement workflows for campaign diagnostics, attribution anomalies, suspicious traffic checks, MMM analysis, and budget-scenario comparison.

05

Implemented scheduled parsers, validation checks, normalized campaign datasets, tracking-link normalization, deduplication, web-to-app joins, and downstream conversion attribution.

06

Developed Go microservices for order processing and payment operations during migration from a monolithic system, with idempotency, retry/backoff, deduplication, caching, messaging, and monitoring.

07

Built backend services and reporting APIs for transportation operations, including delay analytics, route monitoring, SQL optimization, and document workflow automation.

WORKING STYLE

How they show up

  • Works hands-on from requirement clarification and technical specification through implementation, validation, release control, and production monitoring.
  • Keeps core business logic in backend services while using workflow automation tools for alerts, review handoffs, digests, and stakeholder notifications.
  • Has worked in teams of different sizes and described experience with differentiated responsibilities and coverage during absences.
  • Uses structured planning considerations for technology choices, including business constraints, team onboarding, technical requirements, ecosystem fit, maintainability, and vendor lock-in.