AI Engineer – Core

This is a fully remote position, open to applicants in Turkey.

📋 Description

• Develop end-to-end, production-grade AI systems, transitioning from prototypes to pipelines to products.

• Take responsibility for the evaluation layer of Hilbert's agents, which includes harnesses, metrics, golden datasets, regression gates, and human-in-the-loop reviews.

• Establish the definition of correctness for multi-step agent trajectories.

• Create regression gates that are executed prior to releases.

• Transform production failures into sustainable test cases and remedies.

• Design and implement agent workflows using frameworks like LangChain, LangGraph, or similar alternatives.

• Architect state memory, routing, tool registries, and recovery paths.

• Oversee systems from the experimentation phase through to production.

• Manage production systems with a focus on tracing, monitoring, latency management, cost-per-task budgeting, and on-call support.

• Diagnose issues such as hallucinations, tool misuse, retrieval misses, and silent degradation.

• Set technical standards, evaluate designs, and establish reusable agent patterns.

• Collaborate closely with the founding team and cross-functional partners.

• Clearly communicate trade-offs, progress, and technical decisions.

• Integrate retrieval strategies that combine RAG, graph-based retrieval, and other methodologies into a cohesive approach.

• Develop resilient agentic workflows that effectively manage edge cases, missing data, human escalation, and failure prevention.

• Connect agents to external platforms and execute real-world workflows including human-in-the-loop checkpoints.

• Execute, learn, and iterate amidst ambiguity.

⛳️ Requirements

• Minimum of 4 years of experience in building production software.

• At least 2 years of experience in developing LLM or agentic systems utilized by actual users in production environments.

• Proficiency with APIs, services, and data infrastructure.

• Experience in taking ownership of code, including tests, CI/CD, and on-call responsibilities.

• Practical experience with LangChain, LangGraph, or comparable agent/orchestration frameworks.

• Background in designing and implementing agent workflows, covering state memory, routing, tool registries, and recovery paths.

• Capability to operate production systems, focusing on tracing, monitoring, latency, cost-per-task budgeting, and on-call duties.

• Proven experience in diagnosing production failures such as hallucination, tool misuse, retrieval misses, and silent degradation.

• Strong grasp of retrieval-augmented generation, especially hybrid and graph retrieval, chunking, embeddings, ranking, and grounding.

• Familiarity with LLM observability tools like Langfuse, OpenTelemetry, or similar.

• Experience with optimizing costs and latency.

• Knowledge of MCP, tool-calling frameworks, structured output, and constrained decoding is a significant advantage.

• Experience at early-stage startups or in rapidly growing environments is a strong plus.

• Excellent technical communication skills, with the ability to articulate decisions to both technical and non-technical stakeholders.

• Capacity to take initiative, work independently, excel in ambiguous situations, and operate at a startup pace.

• Availability to work with at least 5 hours of overlap with the PST timezone.

🏝️ Benefits

• Competitive salary plus an equity package, aligned with experience.

• Equity package.

• Performance-based bonuses linked to project milestones and customer impact.

• Opportunities for professional growth and increasing responsibilities.

• Genuine ownership of your impact.

• An inclusive culture and equal employment opportunity.