Generative AI Engineer, Afficiency

• This role is ideal for someone who is based in the NYC Metro Area, passionate about building real-world GenAI applications and bringing them into production, while continuously improving performance, reliability, and user outcomes
• Education: Master's degree or equivalent experience required
• 3+ years in software engineering, data engineering, ML engineering, or applied AI, including recent GenAI delivery in production
• Demonstrated expertise in RAG system design and optimization, including:
• Chunking + metadata enrichment, hybrid search, re-ranking, retrieval evaluation
• Grounding/citations and hallucination mitigation patterns
• Strong Python and backend engineering skills (FastAPI/Flask), plus strong SQL
• Experience working in regulated or security-conscious environments, with knowledge of:
• Access controls/entitlements, data privacy, logging/audit trails, secure SDLC practices
• Proven ability to work effectively as an IC consultant:
• Communicate architecture decisions clearly
• Influence cross-functional stakeholders without direct authority produce high-quality documentation and handoff materials

Desirable

• As a Generative AI Engineer at Afficiency, you will be responsible for designing, developing and deploying Generative AI solutions that enhance our core product platforms and client implementations
• You will work closely with engineering, data science, and infrastructure teams to build scalable AI-driven applications using Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), model fine-tuning, and reinforcement learning approaches
• Deliver GenAI solutions end-to-end
• Own technical design and implementation of GenAI applications from discovery through production handoff
• Build APIs/services that integrate with enterprise systems and analytics platforms
• Implement enterprise-grade RAG
• Design ingestion pipelines for internal content (PDFs, policies, research, dashboards, ticketing, wikis)
• Build retrieval systems with hybrid search, filtering, re-ranking, query rewriting, and context optimization
• Implement permission-aware retrieval aligned to entitlements and data access policies
• Establish evaluation and quality controls
• Define metrics for retrieval quality and answer grounding (faithfulness, citation accuracy, coverage)
• Create golden datasets, regression tests, and automated evaluation harnesses
• Operationalize GenAI (LLMOps)
• Instrument observability (latency, cost, token usage, error rates) and implement safe rollout patterns
• Implement caching, rate limiting, fallbacks, and incident-ready operational practices
• Partner across teams to land solutions
• Collaborate with business owners to translate requirements into workable designs
• Work with Security/Compliance to embed guardrails, auditability, and privacy controls
• Provide clear documentation and implementation of playbooks to enable internal teams' post-engagement
• Tools & Technologies
• LLM frameworks: LangChain, LlamaIndex, Semantic Kernel (optional)
• Vector/hybrid search: Open to different skillsets
• Data: (Snowflake/Databricks/warehouse), event pipelines, document stores
• Observability: logging/tracing/metrics, dashboards, alerting