Generative AI Engineer
sia-partners
Job Description
You are part of a cross-functional consulting team that drives the adoption of Generative AI in every imaginable sector, working step-by-step with customers to understand business requirements to design then build bespoke GenAI solutions.
- Build applications powered by LLMs (OpenAI, Claude, Mistral, etc.) using LangChain, LlamaIndex, and related GenAI frameworks.
- Implement RAG pipelines with vector DBs (Pinecone, FAISS, pgvector, ChromaDB) for grounding LLM responses with internal knowledge
- Develop multimodal AI solutions (text, audio, image) and build autonomous agents where relevant.
- Drive MLOps excellence: CI/CD (ML pipelines), drift detection, canary releases, retraining schedules.
- Design robust and reusable prompt templates using CoT, ReAct, Graph-of-Thought, and Agent flows.
- Continuously improve model reliability, relevance, and UX by tuning prompt flows
- Deploy GenAI models on AWS/GCP/Azure using services like SageMaker, Bedrock, Vertex AI
- Ensure performance observability, security guardrails, and compliance (GDPR, Responsible AI)
- Work with DevOps teams to integrate GenAI solutions into microservices and APIs (FastAPI/Flask)
- Benchmark open-source and commercial LLMs for use-case fit and cost-performance tradeoffs
- Evaluate fine-tuning strategies (PEFT, LoRA, RLHF) where applicable for proprietary use cases
- Support solution architects and cross-functional teams in delivering PoCs and enterprise-grade rollouts
- Document frameworks, best practices, risks, and learnings for future scaling
Qualifications
- Bachelor’s/master's degree in computer science, AI, or a related field.
- 5+ years of experience in NLP/ML/AI with at least 3 year hands-on in GenAI.
- Strong coding skills in Python with frameworks like PyTorch, Hugging Face, LangChain, and LlamaIndex.
- Proven experience with cloud-based AI services (AWS/GCP/Azure) and APIs (OpenAI, Anthropic, Hugging Face).
- Experience with vector databases: Qdrant, pgvector, Pinecone, FAISS, Milvus, or Weaviate.
- Familiarity with prompt engineering, transformer architectures, and embedding techniques.
- Excellent communication skills, with the ability to convey complex technical concepts to both highly technical and also non-technical stakeholders.
- Sharp problem-solving skills.
- Ability to collaborate with diverse teams.