AI Software Engineer — LLM, RAG, Internal AI Products

Bluepes is looking for an AI Software Engineer to help us build practical AI-powered products, workflows, and internal systems.

When applying, please mention your current location and when you would be available to start after receiving an offer.

This is a hands-on engineering role. You will work on internal products and tools that use LLMs, RAG, structured outputs, evaluation logic, integrations, and automation workflows. The first key focus will be a new in-house AI product that works with complex knowledge and content workflows, where AI outputs need to be measurable, traceable, and reliable — not just “generated.”

This role is not about training machine learning models. We are looking for an AI-literate software engineer who can build with LLMs, design reliable workflows around them, understand business needs, and turn written specifications into working systems.

Python experience is a strong advantage, but it is not a strict requirement. We are open to candidates whose main stack is different, as long as they have solid software engineering experience and are comfortable working with Python-based tools or willing to work in this direction.

You will work directly with Bluepes management, product stakeholders, and developers. We expect the person to be independent, careful with details, comfortable with written specifications, and proactive in suggesting better technical solutions.

What you will do

• Build and improve AI-powered internal products and workflows for Bluepes.
• Work on LLM-based applications, RAG systems, AI assistants, workflow agents, knowledge bases, CMS/CRM-related tools, reporting, and operational automation.
• Integrate LLM providers such as OpenAI, Anthropic, Google, or local/open-source model endpoints.
• Design reliable structured outputs, including JSON schemas, validation, retries, error handling, and fallback logic.
• Build RAG workflows: document processing, chunking, embeddings, vector search, keyword search, source tracing, and retrieval quality checks.
• Design evaluation logic for AI-generated outputs: quality criteria, scoring, consistency checks, confidence signals, and human review points.
• Integrate AI workflows with internal systems, APIs, databases, CMS, CRM, Git repositories, automation tools, and business processes.
• Write clean, maintainable code with tests and clear documentation.
• Work from written specifications, identify unclear requirements, ask good questions, and keep implementation aligned with the product logic.
• Help us test new AI tools and approaches, while staying focused on practical, reliable systems rather than hype.

Required skills

• 4+ years of commercial software development experience.
• Strong hands-on software engineering experience in backend, full-stack, product engineering, or internal tools development.
• Practical experience with at least one modern programming stack, such as Python, JavaScript/TypeScript, Java or similar.
• Practical experience building with LLM APIs or AI-assisted applications.
• Good understanding of RAG, embeddings, vector search, semantic search, prompting, and context management.
• Experience with structured LLM outputs, validation, retries, and handling unreliable or malformed model responses.
• Solid understanding of APIs, HTTP, SQL/databases, file handling, integrations, Git, testing, and software delivery basics.
• Ability to work with Python-based tools, scripts, or services, even if Python is not your main commercial stack.
• Ability to build or adjust simple frontend/UI flows when needed — JavaScript, HTML/CSS, browser APIs, fetch, DOM.
• Ability to read and work from detailed written specifications.
• Strong ownership: you can move independently, raise risks early, and propose solutions instead of waiting for fully prepared tasks.
• Clear written communication in English.
• English level: Upper-Intermediate or higher.

Nice to have

• Strong Python experience.
• Experience with LangChain, LangGraph, LlamaIndex, Semantic Kernel, or similar AI engineering frameworks.
• Experience with AI evaluation tools or practices: RAGAS, LangSmith, Langfuse, tracing, prompt/version control, graders, observability.
• Experience with vector databases or search tools such as sqlite-vec, FAISS, pgvector, Elasticsearch, OpenSearch, or BM25/hybrid search.
• Experience with agentic workflows, tool calling, multi-step LLM orchestration, human-in-the-loop workflows, or autonomous/semi-autonomous agents.
• Experience with CMS, CRM, knowledge management systems, reporting tools, or internal business automation.
• Experience with workflow automation tools such as n8n, Make, Zapier, or similar.
• Experience with Docker, GitHub Actions, CI/CD, AWS, Azure, GCP, or serverless tools.
• Experience with local models, Ollama, or OpenAI-compatible local endpoints.
• Security mindset around AI systems: prompt injection, data leakage, access control, safe file handling, and safe tool execution.
• Experience building products from an early stage.

What we offer

• Full-time remote position.
• Long-term cooperation with an established software company.
• Work on internal AI-powered products and automation projects, not short-term client experiments.
• Space to propose ideas, test new tools, and influence technical decisions.
• Practical AI engineering work: LLMs, RAG, evaluation, integrations, workflow automation, and internal product development.
• 18 paid vacation days per year.

Recruitment process

Our recruitment process is simple and transparent:

1. Intro call with the recruiter — a short conversation to check general fit, cooperation terms, availability, and English level.
2. Technical interview — a technical conversation fully in English.
3. Paid test task — a small practical task for selected candidates, paid regardless of the final decision.
4. Offer call — if everything looks good on both sides, we will have a final call to discuss the offer and next steps.

*Due to the high number of applications, we may not be able to reply to every candidate individually. If you do not hear from us after applying, it means we will not be moving forward with your application at this stage. Thank you for your understanding.