LLM Application Engineer: ChatGPT & AI Pipelines

Devitechs; Texas; Vor Ort; USD 80.000 - 120.000; Vollzeit

Zusammenfassung

Devitechs is looking for a professional to design and develop end-to-end LLM-powered applications using cutting-edge technologies like GPT-4 and GPT-5. This role focuses on integrating ChatGPT with backend systems and building scalable AI pipelines for diverse applications.

Candidates will implement various features including moderation and safety filters, while ensuring secure handling of sensitive data, and collaborate closely with product and design teams.

Aufgaben

• Design and develop end-to-end LLM-powered applications using GPT-4, GPT-5 or custom models.
• Integrate ChatGPT with backend systems through REST APIs, WebSockets, and cloud functions.
• Build scalable AI pipelines for summarization, content generation, Q&A, and automation tasks.
• Fine-tune models using available datasets to improve accuracy for specific business domains.
• Implement guardrails, moderation rules, and safety filters to control model output.
• Build vector search systems using FAISS, Pinecone, Weaviate, or Milvus.
• Work with RAG (Retrieval Augmented Generation) frameworks for grounded responses.
• Implement caching, batching, and optimization for cost reduction and performance.
• Develop structured workflows for approvals, escalations, and error handling in LLM interactions.
• Monitor LLM performance through logs, metrics, and analytics dashboards.
• Test model outputs for correctness, consistency, and hallucination risks.
• Collaborate with product and design teams to create LLM-driven UX flows.
• Maintain secure handling of sensitive data used in model queries.
• Implement multi-agent systems for complex AI interactions and decision-making.
• Automate repetitive business tasks using ChatGPT-powered agents.
• Document data flows, architecture, model behaviors, and system dependencies.
• Troubleshoot model-related issues including context loss and misclassification.
• Scale LLM applications using Kubernetes, serverless functions, and cloud orchestration.
• Stay updated with the latest GPT model releases, research papers, and API features.
• Provide internal training on LLM capabilities, limitations, and best practices.

Jobbeschreibung

✔️ Design and develop end-to-end LLM-powered applications using GPT-4, GPT-5 or custom models.

✔️ Integrate ChatGPT with backend systems through REST APIs, WebSockets, and cloud functions.

✔️ Build scalable AI pipelines for summarization, content generation, Q&A, and automation tasks.

✔️ Fine-tune models using available datasets to improve accuracy for specific business domains.

✔️ Implement guardrails, moderation rules, and safety filters to control model output.

✔️ Build vector search systems using FAISS, Pinecone, Weaviate, or Milvus.

✔️ Work with RAG (Retrieval Augmented Generation) frameworks for grounded responses.

✔️ Implement caching, batching, and optimization for cost reduction and performance.

✔️ Develop structured workflows for approvals, escalations, and error handling in LLM interactions.

✔️ Monitor LLM performance through logs, metrics, and analytics dashboards.

✔️ Test model outputs for correctness, consistency, and hallucination risks.

✔️ Collaborate with product and design teams to create LLM-driven UX flows.

✔️ Maintain secure handling of sensitive data used in model queries.

✔️ Implement multi-agent systems for complex AI interactions and decision-making.

✔️ Automate repetitive business tasks using ChatGPT-powered agents.

✔️ Document data flows, architecture, model behaviors, and system dependencies.

✔️ Troubleshoot model-related issues including context loss and misclassification.

✔️ Scale LLM applications using Kubernetes, serverless functions, and cloud orchestration.

✔️ Stay updated with the latest GPT model releases, research papers, and API features.

✔️ Provide internal training on LLM capabilities, limitations, and best practices.