Ілля

Illia Ivashko Python Backend & AI Developer
[відкрити контакти](див. вище в блоці «контактна інформація») Lviv linkedin github @illia_python

PROFESSIONAL SUMMARY
Python Backend Developer focusing on AI-driven applications, LLMOps, and microservices architecture. Proficient in building high-
performance REST APIs with FastAPI and Django Ninja, orchestrating multi-agent workflows (LangGraph, Function Calling), and
optimizing LLM latency and token costs. Experienced in fintech payment integrations, multi-tenant B2B architectures, and resilient
data extraction pipelines. Passionate about backend performance tuning, reliable asynchronous processing, and delivering scalable,
business-focused solutions.
WORK EXPERIENCE
Python Backend & AI Developer, UK Travel Tech Startup (NDA) 11/2024 – Present
Project Overview: Developed an AI-powered conversational microservice integrated with the core platform, enabling users to
seamlessly query travel data and execute transactional workflows via chat.
Tech Stack: Python (Asyncio), FastAPI, SQLAlchemy, Alembic, LangChain, LangGraph, LiteLLM, Anthropic SDK, OpenAI API, Tavily,
PostgreSQL (pgvector), Redis, AWS (S3, EC2), WebSockets/SSE, Langfuse, LangSmith, Prometheus, Grafana, Kubernetes, Docker,
GitLab CI, uv, Pytest.
• Agent Orchestration: Implemented LangGraph state machines with dynamic Function Calling to query core backend APIs and Tavily for
real-time web retrieval.
• Semantic Caching: Engineered a Redis prompt-caching layer using vector embeddings; cut external LLM API calls by ~45% and
reduced p95 latency from 1.8s to 650ms.
• Observability & CI/CD: Built an end-to-end tracing and prompt evaluation pipeline using Langfuse and LangSmith; maintained 90%+
test coverage via Pytest.
Python Backend Developer, Centum-D 03/2024 – 10/2024
Project Overview: Developed a multi-tenant backend for a B2B project planning platform, enabling users to architect projects with AI
assistance and processing secure financial transactions.
Tech Stack: Python, SQL, Django, Django Ninja, Pydantic v2, PostgreSQL, Qdrant, RabbitMQ, Celery, REST API, GraphQL, KrakenD,
Auth0, OAuth2, Rapyd API, GitHub Actions, Docker Compose, Poetry.
• Type-Safe APIs: Designed core REST and GraphQL services using Django Ninja and Pydantic v2, enforcing strict multi-tenant data
segregation.
• Fintech Payment Integration: Integrated Rapyd API with asynchronous Celery & RabbitMQ webhook handlers; implemented
idempotency locks eliminating duplicate charges.
Python Developer, Appexoft 09/2022 – 02/2024
Project Overview: Worked on diverse client projects, ranging from full-stack e-commerce development to high-volume data aggregation
and legacy system maintenance.
Tech Stack: Python, SQL, Django, DRF, PostgreSQL, MySQL, MongoDB, Kafka, Stripe API, GCP, Playwright, Selenium, cloudscraper,
Scrapy, BeautifulSoup4, Unittest, Pytest.
• E-commerce & Payments: Built custom e-commerce backend using Django/DRF with Stripe API checkout; optimized database indexes
speeding queries by 40%.
• Distributed Extraction: Engineered GCP-deployed scraping pipelines with Playwright/Selenium, streaming clean aggregated data via
Kafka into MongoDB.
TECHNICAL SKILLS
Languages AI & LLMOps Frameworks & Libraries
Python (Asyncio, OOP), SQL, Bash. LangChain, LangGraph, LiteLLM, Anthropic FastAPI, Django, Django Ninja, Django REST
Architecture & Protocols SDK, OpenAI API, Ollama, LangSmith, Framework (DRF), Pydantic v2, SQLAlchemy,
REST API, GraphQL, WebSockets, Server- Langfuse, Tavily. Alembic.
Sent Events (SSE), Webhooks, KrakenD Message Brokers & Async Databases
Gateway. RabbitMQ, Celery, Apache Kafka (Basic). PostgreSQL, MySQL, Redis, MongoDB.
Vector Search & DBs DevOps & Cloud Observability
Qdrant, pgvector (PostgreSQL). AWS (S3, EC2), GCP, Docker, Kubernetes, Prometheus, Grafana, Jaeger.
Testing & Scraping CI/CD (GitHub Actions, GitLab CI), Nginx.
Pytest, Unittest, Playwright, Selenium, Tools & Environment
BeautifulSoup4, Scrapy, cloudscraper. Git, Poetry, uv, Auth0, OAuth2, Stripe API,
Rapyd API, Postman, Swagger/OpenAPI.
LANGUAGES
Ukrainian — Native | English — B2 (Upper-Intermediate)