AI Engineer

NevkaSystems; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

NevkaSystems is looking for an engineer to build AI features that go beyond demos, including conversational agents, knowledge assistants and intelligent workflows. You will connect models to business systems and measure performance in real-world use.

You will design prompts, pipelines and evaluations, aiming to improve latency, security and user experience while turning prototypes into production systems. Strong Python or TypeScript skills and API experience are highly valued.

Qualifications

• Practical experience building AI applications with large language models and external APIs.
• Strong Python or TypeScript skills and experience developing backend services.
• Understanding tool calling, structured outputs, retrieval-augmented generation and embeddings.
• Experience evaluating model outputs with realistic test cases and investigating inconsistencies.
• Working knowledge of databases, authentication, Git, automated testing and deployment workflows.
• Ability to explain engineering trade-offs and collaborate clearly in a remote team.
• Experience with real-time voice, speech systems or multilingual agents is a plus.

Responsibilities

• Build and ship AI features, including conversational agents, voice experiences and workflow automation.
• Connect agents to business APIs and tools with clear permissions, validation and human handoffs where needed.
• Develop retrieval and knowledge pipelines, including document ingestion, search and source-grounded responses.
• Design prompts, conversation flows and evaluations that test task completion, factual accuracy and failure handling.
• Improve response speed, streaming, reliability and cost through measurement and controlled experiments.
• Implement logging, automated tests and protections for sensitive data, prompt injection and unintended tool actions.
• Work with product, design and engineering colleagues to turn prototypes into maintainable production systems.

Skills

Python

TypeScript

Backend development

LLM tooling

Testing & deployment

Remote-team collaboration

Communication

Tools

Git

Databases

APIs

Authentication

Logging

Job description

Build AI products that work beyond the demo. You will develop conversational agents, knowledge assistants and intelligent workflows, connecting models to business systems and measuring how reliably they perform. We are looking for an engineer who combines strong software fundamentals with a practical approach to model quality, latency, security and user experience.

Responsibilities

• Build and ship AI features, including conversational agents, voice experiences and workflow automation.
• Connect agents to business APIs and tools with clear permissions, validation and human handoffs where needed.
• Develop retrieval and knowledge pipelines, including document ingestion, search and source-grounded responses.
• Design prompts, conversation flows and evaluations that test task completion, factual accuracy and failure handling.
• Improve response speed, streaming, reliability and cost through measurement and controlled experiments.
• Implement logging, automated tests and protections for sensitive data, prompt injection and unintended tool actions.
• Work with product, design and engineering colleagues to turn prototypes into maintainable production systems.

What you bring

• Practical experience building AI applications with large language models and external APIs.
• Strong Python or TypeScript skills and experience developing backend services.
• Understanding of tool calling, structured outputs, retrieval-augmented generation and embeddings.
• Experience evaluating model outputs using realistic test cases and investigating inconsistent behaviour.
• Working knowledge of databases, authentication, Git, automated testing and deployment workflows.
• Ability to explain engineering trade-offs and collaborate clearly in a remote team.
• Experience with real-time voice, speech systems or multilingual agents is a plus. Please share relevant projects or a GitHub link if available.

Fully remote. Applicant location: Worldwide.