Skip to content
EP

evelina penkova

@etoevelina

AI engineer building production LLM, RAG, and FastAPI systems, focused on reliability, evaluation, performance, and meaningful user impact

I’m an AI engineer with two years of production experience, building reliable LLM applications, RAG pipelines, FastAPI services, evaluation systems, and scalable cloud infrastructure focused on measurable product and user impact.

Career

Experience: 2 years

Experience

AI Engineer

KuickTech

Owned production LLM and retrieval work across four products, with LikePost cost reduced about 10x and latency reduced from about 2 minutes to 10 to 30 seconds.

• LikePost, multi-service AI content platform. Sole AI engineer owning the full AI layer of a production social platform with 5+ FastAPI microservices serving tens of thousands of users, including architecture, provider integration, LLM orchestration, retrieval, prompts, evaluation and deployment. • Built a 6 to 8 stage async conversational recommendation pipeline with query expansion, topic-floor gate, hybrid retrieval, LLM location and relevance judges, grounded answers and quick replies, with realtime delivery over Redis pub/sub.

Founder / AI Engineer

SkinX

AI-Powered Skin Analysis Platform

Founded and built a live skin-analysis product, owning its computer vision pipeline and product lifecycle.

• Built, launched and continue to develop a production AI product that is live and operating with real users. • Implemented the end-to-end computer-vision pipeline for skin image analysis, including lesion detection, segmentation, classification and explainability, with structured application-ready outputs. • Evolved the product beyond its original hackathon-winning prototype.

Education

B.Sc., Software Engineering

Astana IT University

2024

Applied Informatics

MGIMO College

2020 — 2024

Selected projects

Smart Scheduling System, Automated Timetabling

Designed a constraint-based scheduling engine to automate teacher, classroom and curriculum planning.

Skills

OpenAIxAI GrokGoogle GeminiPerplexityAI/ML PipelinesRAGLangChainPrompt EngineeringLLM OrchestrationMulti-Provider Routing/FallbackPrompt/Model EvaluationLLM-as-JudgeEmbeddingsVector/Semantic SearchOCRWhisperMultimodal AIPythonFastAPIPostgreSQL