Senior Data Scientist

Responsibilities

• Design and develop the core agent layer for orchestrating interactions with LLMs.
• Build and maintain conversational logic, including state machines, agent workflows, and orchestration pipelines.
• Control LLM behavior using prompt design, structured outputs, and deterministic flows.
• Manage conversational context, memory, history, token limits, and degradation strategies.
• Ensure reliability and predictability of non-deterministic models.
• Implement resilient integrations with LLM providers using timeouts, retries, fallbacks, and multi-provider strategies.
• Optimize latency and cost through streaming, batching, caching, and token efficiency.
• Debug production issues including inconsistent outputs, race conditions, and state loss.
• Contribute to architecture with clear boundaries between agents, backend, and real-time components.
• Build observability around LLM pipelines with prompt and response logging, tracing, and quality metrics.

Conditions

Remote work is possible. Payment terms are open to discussion from 3500 €. The company is ready to discuss countries other than Serbia and Armenia.

About the product

FlameTree is building a platform for creating AI agents that help businesses scale customer support, lead follow-up, and sales across inbound and outbound communication channels. The platform works with knowledge bases, communicates in real time, supports 150+ languages, and integrates with WhatsApp, email, and web applications.

What makes this role interesting

• Work on the core intelligence layer of the product rather than only integrations.
• Solve production challenges involving high load, low latency, and reliability.
• Have a direct impact on system architecture and technical decisions.
• Work in a fast execution cycle with minimal bureaucracy.
• Use an engineering-driven approach focused on reliability, control, and metrics.
• Join an engineering team focused on building production systems rather than prototypes.