AI Integration Engineer

The AI Integration Engineer builds the connective tissue between LLM APIs, existing product systems, and end users. They are not training models. They are integrating, orchestrating, and deploying AI capabilities into production software, reliably, securely, and at scale.

Key Responsibilities:

● Integrate LLM APIs (Anthropic Claude, OpenAI GPT-4o, AWS Bedrock, Google

Gemini) into backend services and user-facing products

● Design and implement RAG pipelines: document ingestion, chunking strategy,

vector store selection, retrieval tuning

● Build agentic workflows using frameworks such as AgentCore, LangChain,

LlamaIndex, or custom orchestration patterns

● Manage prompt engineering, prompt versioning, and prompt evaluation

frameworks

● Implement guardrails for LLM outputs: validation, content filtering, fallback logic

● Monitor AI system performance: latency, cost-per-query, accuracy drift, token

usage

● Collaborate with frontend engineers to surface AI capabilities in product UIs

● Own the AI integration layer in the SDLC, from spec to CI/CD to production

observability

Education and Experience:

• A degree (Masters preferred) in Computer Science, Engineering, Artificial Intelligence, Data Science, Applied Mathematics, or related fields.

• Minimum 5+ years of hands-on experience in engineering.

Technical Skills:

• 4+ years backend or full stack experience; strong API design and consumption skills

• Proven experience integrating LLM APIs (any major provider) into production applications, not just prototypes

• Proficiency in Python and/or TypeScript/

• Hands-on experience with RAG vector databases Pinecone, Weaviate, pgvector etc), embedding models, chunking strategies.

• Understanding of prompt engineering: system prompts, few-shot examples, chain-of-thought, structured output

• Solid grasp of API security, rate limiting, cost management for LLM-based services

• Experience with AWS or another major cloud platform

• Desirable Skills:

• Experience with agentic frameworks: AgentCore, LangChain, LlamaIndex,

CrewAI, AutoGen, or similar

• Familiarity with multi-modal AI (vision, audio) or function calling / tool use