Python Developer

Quick Facts

• Role: Python Developer (AI infrastructure for LLM context management)

Description

You’ll build and improve systems that decide what content enters an LLM context window—its form, and why. The work spans text processing, information retrieval, and agent architectures, with a strong focus on measurable quality and cost through benchmark-based evaluation.

Responsibilities

• Develop text and context processing pipelines (segmentation, deduplication, scoring, selection for the LLM context window)

• Work on information retrieval components (indexing, ranking, embeddings, search quality evaluation)

• Build and test agent components (tool-use loops, memory/context management, orchestration)

• Build and run benchmark evaluations; measure impact of each change on quality and cost

• Rapidly prototype and evolve prototypes into maintainable production systems

• Iterate in short cycles with direct feedback from a technical lead

Requirements

• Strong Python skills, including data structures, complexity analysis, profiling, and writing readable testable code

• Practical text processing experience: tokenization, chunking, embeddings, and handling large text volumes

• Algorithmic thinking for designing and justifying segmentation, deduplication, or scoring approaches

• Comfort working with Linux and Git

• Knowledge and ability to use tokenizers

• Experience with at least one: LLM agent architecture (tool use, planning, context management), large-scale deduplication (e.g., MinHash/SimHash) or ranking systems, a real working RAG/agent project used by others, or open-source contributions in NLP/LLM tooling

• Basic information retrieval knowledge: BM25, vector search, and quality metrics such as precision/recall and NDCG

Benefits

• Salary range: 12,000–25,000 PLN (contracting/B2B; depends on the technical interview outcome)

• Work on AI infrastructure with measurable results on public benchmarks

• Close collaboration with a technical lead (fast growth, no intermediaries)

• Flexible working hours; office in Warsaw (Połczyńska 112A)

• Access to paid AI tools (Claude, Copilot, Codex, ChatGPT, Gemini, etc.)

• Modern infrastructure, including B200/B300 clusters

• Training budget and conference participation

• Company hackathons