AI Data Scientist

Amtex Enterprises Inc; Princeton, Northern (NJ, KY); Hybrid; USD 120,000 - 170,000; Full time

Job summary

Amtex Enterprises Inc is seeking an AI Data Scientist to lead the development of retrieval-augmented AI solutions. The role covers the full AI lifecycle from data ingestion to deployment, monitoring, and support in a remote-capable environment.

The ideal candidate has deep expertise in RAG architectures, embeddings, vector databases, and enterprise-grade ML with explainability and governance considerations.

Qualifications

• Experience building production-grade RAG systems.
• Strong ML/statistics background with explainability.
• Proficiency with end-to-end AI lifecycle delivery.

Responsibilities

• Lead ingestion, retrieval, modeling, APIs, tests, deployment.
• Develop explainable, auditable AI in regulated environments.
• Evaluate retrieval quality and grounding with LLMs.

Skills

Python

Pandas

NumPy

Scikit-learn

PyTorch

RAG

Embeddings

Vector databases

Explainability

MLOps

Tools

Git

CI/CD

Docker

Podman

API development

Databricks

Azure

pgvector

Qdrant

Job description

Job Title : AI Data Scientist
Duration: 6+ Months
Location: Can be fully remote. Candidates in EST or CST highly preferred.

We are looking for a senior practitioner who has successfully built and deployed at least one production-grade Retrieval Augmented Generation (RAG) solution. The ideal candidate should be capable of working across the full AI lifecycle, including ingestion, retrieval, modeling, APIs, testing, deployment, monitoring, and operational support.
The candidate should also have strong expertise in classical machine learning and statistics and be experienced in delivering explainable, auditable AI solutions within regulated, human-in-the-loop environments.

Here are additional details

Machine Learning & Applied AI

• Strong background in supervised and unsupervised learning, including:
  • Classification
  • Ranking
  • Clustering
  • Anomaly detection
  • Predictive modeling
• Experience selecting evaluation metrics and designing representative test datasets
• Hands-on experience with explainability techniques such as SHAP, LIME, and feature importance analysis
• Experience designing human-in-the-loop AI systems with review, escalation, override, and feedback mechanisms

Generative AI & Retrieval

• Production experience with:
  • RAG architectures
  • Embeddings
  • Semantic search
  • Re-ranking
  • Prompt engineering
  • Vector databases
• Experience building enterprise copilots, assistants, or document-grounded decision-support systems
• Ability to evaluate retrieval quality, grounding, hallucination risk, answer quality, and failure modes
• Experience with open-source or locally hosted LLMs preferred

Python & Software Engineering

• Advanced Python skills, including:
  • Pandas
  • NumPy
  • Scikit-learn
  • PyTorch (or similar deep learning framework)
• Ability to develop maintainable, production-quality code rather than notebook-only solutions
• Experience with:
  • Git and code reviews
  • Unit and integration testing
  • CI/CD pipelines
  • API development
  • Docker or Podman
• Experience with agentic development is a plus
• Comfortable working in both local and cloud environments

Data Engineering & Integration

• Strong SQL and ETL/ELT development skills
• Experience integrating:
  • REST APIs
  • Enterprise document repositories
  • Workflow systems
  • Batch and incremental data pipelines
• Experience with PostgreSQL, SQL Server, data lake architectures, data lineage, and data quality controls
• Ability to design resilient ingestion pipelines for documents, metadata, attachments, and changing source-system records

Enterprise AI, Governance & Operations

• Experience implementing:
  • Audit trails
  • Evidence and reasoning logs
  • Monitoring and telemetry
  • Operational diagnostics
• Understanding of:
  • AI governance
  • Security and privacy requirements
  • Model validation
  • Access controls
  • Traceability and records retention
• Ability to design AI systems that support accountable human decision-making
• Experience communicating confidence levels, limitations, failure modes, and operational boundaries

Preferred Technology Stack

• Azure
• Databricks
• PostgreSQL with pgvector
• Qdrant or similar vector databases
• Ollama or other local/on-prem model serving platforms
• OCR and document processing solutions
• Knowledge graphs
• Semantic evaluation frameworks
• Agent orchestration tools
• Experience deploying AI solutions in restricted, on-premises, or hybrid environments