- Company Name
- Coris
- Job Title
- AI Engineer
- Job Description
-
Job title: AI Engineer
Role Summary: Design, fine‑tune, and deploy large and small language models for fraud detection and risk mitigation, while building scalable backend infrastructure that integrates model outputs into production systems and APIs.
Expactations: Deliver low‑latency, high‑accuracy fraud‑prediction models that adapt to evolving adversaries, maintain cost efficiency, and meet compliance standards. Own end‑to‑end model life cycle, from data preparation to inference serving, and collaborate across data, engineering, and product teams to iterate quickly and measure impact rigorously.
Key Responsibilities:
• Fine‑tune, distill, and quantize LLMs/SLMs (LoRA/PEFT, 8‑bit, etc.) for fraud use cases such as entity resolution, anomaly detection, and customer‑communication classification.
• Optimize inference pipelines using vLLM, TensorRT, and other serving frameworks to meet millisecond latency targets and scale to hundreds of millions of events per month.
• Build end‑to‑end training and evaluation pipelines that balance recall and precision, create golden and adversarial datasets, and implement online/offline evaluation harnesses.
• Engineer feature pipelines extracting latent signals and non‑obvious features for fraud models.
• Architect Python/Django services that expose model predictions via customer APIs and integrate with existing fraud workflows.
• Model fraud/risk data in Postgres, ensuring query and aggregation performance at billions of rows.
• Develop and maintain data ingestion pipelines from payment processors (Stripe, Adyen, etc.) with near‑real‑time throughput.
• Implement observability—logs, metrics, drift detection—to monitor model performance and detect changes in fraud tactics.
Required Skills:
• 3+ years building production systems in Python/Django with Postgres.
• Hands‑on experience fine‑tuning and optimizing LLMs/SLMs for adversarial or anomaly‑detection tasks.
• Proven ability to reduce inference latency and cost without degrading accuracy.
• Strong command of PyTorch/TensorFlow, profiling tools, and modern serving stacks (vLLM, TensorRT, Triton).
• Experience designing and scaling feature pipelines, feature stores, and online learning mechanisms.
• Familiarity with PII, KYC/AML, and financial compliance requirements.
• Practical, experimental mindset: ship fast, measure rigorously, iterate.
Required Education & Certifications:
• Bachelor’s degree in Computer Science, Data Science, Engineering, or related field (or equivalent experience).
• Relevant certifications in machine learning, data engineering, or cloud platforms are a plus.
San francisco, United states
Hybrid
18-10-2025