4 months ago
Kaunas, Lithuania or Vilnius, LithuaniaStaff+
Responsibilities
- Shape AI/ML technical direction, research strategy, proof-of-concept validation, and production architecture.
- Lead hands-on exploration and production implementation of context management, memory, retrieval, agentic loops, A2A, frameworks, and tooling.
- Contribute across product discovery and delivery as both a hands-on builder and technical thought leader.
- Own the LLM evaluation framework, including criteria, assessments, and decisions about performance, cost, and reliability.
- Act as an internal and external AI evangelist and influence the team’s technical direction.
Requirements
- Proven experience driving AI/ML initiatives end-to-end in production and owning outcomes.
- Expertise designing, executing, and scaling production-grade AI systems with hands-on implementation ability.
- Deep knowledge of AI developer frameworks and tooling, large language models, embeddings, multimodality, and AI architectures.
- Advanced expertise in context engineering and deployed evaluations.
- Strong software engineering fundamentals, microservice architectures, and taking proofs of concept through production.
- Preferred experience designing and shipping complex agentic systems at scale.
- Preferred expertise in data science methodologies, ML algorithms, and natural language processing.
- Preferred experience with backend engineering, API design, scalable systems, Kubernetes, and end-to-end feature ownership across engineering, product, sales, and marketing.
Benefits
- Professional development through events, online training, conferences, and books.
- Health benefits including private health insurance, workouts, sports access, and consultations.
- Team-building events, parties, food coupons, and gifts.
- Additional vacation days based on tenure.
- Additional paid leave for illness or special occasions.
- Flexible working time arrangement.
Tech Stack
Categories
About nexos.ai
Nexos.ai builds an AI Gateway and Workspace that give companies a single, EU-hosted API and no-code layer to access, route, and monitor 200+ models (Claude, GPT, Gemini, Mistral). The platform adds spend tracking, rate limiting, failover, and compliance controls, letting teams swap providers without rewriting code. Founded in 2024 and headquartered in Vilnius, the privately held company sells usage-based SaaS to engineering and operations teams deploying LLM-powered workflows.