Antares Solutions

AI Engineer

Chicago, USA

No sponsorship$175k-$240kDetected 91 days ago

PythonDatabricksVector DatabasesAzureDockerTerraformCI/CDGraphQLRESTMachine LearningSparkData EngineeringLLMsRAGMLOpsA/B TestingUnityCommunication

About the role

Job Description Antares Capital is seeking an AI Engineer to join our Data & Analytics Technology team.
Base Salary Range $175,000 - $240,000 To learn more, visit www.antares.com.
Antares is an Equal Opportunity Employer.

Design and implement robust RAG pipelines integrating domain datasets, embeddings, and retrieval strategies to deliver accurate, auditable responses.
Lead the evaluation and integration of vector databases (e.g., FAISS, Pinecone, Milvus) and tune indexing/embedding strategies for performance and relevance.
Architect and orchestrate combinations of LLMs and tools (routing, ensemble prompts, function-calling, guardrails) to optimize quality, latency, and cost.
Partner with data and platform teams to establish and evolve a semantic layer that aligns data products with business entities, definitions, and policies
Contribute to and extend the AI reference architecture emphasizing modular services, clear interfaces, observability, and change-tolerant design.
Develop secure data access patterns (role-based permissions, PII minimization) and implement content filtering, redaction, and safety controls.
Build evaluation frameworks (automated tests, offline/online metrics, human-in-the-loop review) and maintain datasets for regression benchmarking.
Implement CI/CD and containerization for AI services
Collaborate with product, data, risk, and security teams to translate business needs into pragmatic AI solutions aligned to industry compliance and model risk management.
Troubleshoot production issues, conduct post-incident reviews, and drive reliability improvements (SLOs, error budgets, resilience testing).

Hands-on expertise with RAG: embedding generation, retrievers, prompt construction, context management, and hallucination mitigation.
ability to tune similarity search (cosine, dot-product) and index parameters.
Must have unrestricted authorization to work in the United States.
Must be willing to comply with pre-employment screening, including but not limited to drug testing, reference verification, and background check.
5+ years of industry experience building and deploying AI/ML applications, including 2+ years with LLM-based systems (preferably in financial services).
Deep understanding of vector databases and embedding frameworks
Proven experience with ontology-driven data modeling (business entities, taxonomies, knowledge graphs, semantic modeling) and mapping from physical schemas to conceptual models. Any experience with 3rd party platform (eg: Palantir/Foundry) implementations is a plus.
Proficiency with big data and machine learning platforms such as Databricks (Spark, Delta Lake, Unity Catalog) and experience operating at scale.
Awareness of financial-industry considerations: data privacy, model risk/governance, auditability, and secure development practices.

Fluency in Python and production-grade services (microservices, REST/GraphQL, event-driven patterns)
strong software engineering fundamentals.
Experience with large-scale cloud data/AI solutions, including Microsoft Fabric (OneLake, Lakehouse, semantic models, pipelines) or equivalent enterprise data/AI fabric, and common cloud services (Azure preferred).
Grounding LLMs with curated, versioned knowledge sources
experience with data pipelines and ETL/ELT concepts.
Strong grasp of evaluation, observability, and MLOps for LLMs (dataset management, A/B testing, drift/quality monitoring, prompt/version governance).
Practical experience with CI/CD, Docker/containers, and infrastructure-as-code (Terraform or equivalent).
Proven experience with ontology-driven data modeling (business entities, taxonomies, knowledge graphs, semantic modeling) and mapping from physical schemas to conceptual models.

Drive an ontology-driven approach: model and map enterprise data to real-world business concepts (e.g., customers, counterparties, facilities, equipment) rather than siloed technical tables
Drive an ontology-driven approach: model and map enterprise data to real-world business concepts (e.g., customers, counterparties, facilities, equipment) rather than siloed technical tables; steward canonical vocabularies, taxonomies, and knowledge graphs.

This listing is sourced directly from Antares Solutions's careers page and normalized into a canonical job model.