Junior–Strong Junior AI Developer

Brainence has spent 9+ years building first-class products — and the teams behind them. Over 50 startups and enterprises have trusted us to launch MVPs, scale engineering teams, and deliver products that perform in the real world. Our playground spans healthcare, edtech, real estate, retail, construction, logistics, insurance, and more.

You’ll touch several exciting projects, but your main playground will be a Ukrainian startup building a next-gen smart calendar. Think notes, tasks, reminders, and a powerful AI assistant all in one seamless experience. You’ll be coding, integrating, and learning every step of the way.

• Agentic systems. Designing and implementing AI agent workflows with LangGraph/LangChain: tool usage, function calling, structured outputs (JSON schemas), memory management, retries, and fallbacks
• Model integration. Connecting with providers (OpenAI, Anthropic, Google Gemini, Mistral, Cohere) and gateways (OpenRouter, LiteLLM). Implementing routing logic to optimize for quality, cost, and latency, as well as building guardrails and fallbacks
• Retrieval-augmented generation (RAG). Contributing to retrieval pipelines: chunking, embeddings, metadata filtering, vector search. Enhancing quality with re-rankers, semantic caching, and prompt/context optimization
• Vector databases. Working with Pinecone, Weaviate, Qdrant, pgvector: managing namespaces, configuring HNSW/IVF indexes, and optimizing storage tiers
• Data ingestion. Building connectors for various data sources (PDFs, Office docs, emails, CSV/JSON, SaaS APIs), handling metadata normalization, PII redaction, and data lineage tracking
• Evaluation and testing. Supporting offline and online evaluations (golden sets, A/B testing), measuring model faithfulness/groundedness, and setting up automated regression gates in CI/CD
• Apps and APIs. Delivering backend services in Python (FastAPI) with streaming (SSE/WebSockets), authentication/RBAC, and multi-tenant support

• 1–2 years of experience in backend or AI application development
• Strong Python skills for building services and pipelines (FastAPI, asyncio, pydantic, pytest)
• Practical experience with LangChain/LangGraph and prompt/tool design (structured outputs, function calling)
• Familiarity with vector databases (Pinecone, Weaviate, Qdrant, pgvector) and embeddings
• Experience integrating LLM providers (OpenAI, Anthropic, Gemini, Mistral, Cohere) or gateways (OpenRouter, LiteLLM)
• Basic knowledge of evaluation frameworks (Ragas, DeepEval) and observability tools (Langfuse, OpenTelemetry)
• English and Ukrainian languages at Upper Intermediate level or higher

Nice to have:

• Any experience with re-rankers, OCR, or improving semantic search
• Experience with Redis/Postgres for queues or caches
• Interest in trying out local AI models (e.g., Ollama)
• Motivation to grow in AI infrastructure/LLMOps (deep knowledge not required)

Long-term project: 1 year or more
Probationary period: Yes, reviewed after three months
Work schedule: Full-time with flexible working hours

• 18 paid vacation days
• 10 paid sick days
• 2 unpaid days off per month
• Work remotely or from our comfortable Lviv office
• Necessary equipment, or $200 equipment compensation per year
• 50% compensation for professional development costs, including courses, webinars, and training programs