AI Machine Learning Engineer

InterConnect Techs has been invited to submit a National Science Foundation Small Business Innovation Research Phase I proposal for the Evidence Completeness Audit Platform, a biomedical informatics platform designed to improve confidence in published clinical evidence.

We are seeking an AI/ML Engineer to design and implement the platform’s core computational pipeline. The system will compare ClinicalTrials.gov registry records with corresponding journal publications to identify and classify reporting discrepancies.

This is a remote, part-time independent contractor position contingent upon NSF SBIR Phase I award. The engagement will include approximately 400 total hours over Months 1 through 11 of the 12-month project period. Work will be milestone-driven, with variable hours averaging approximately nine hours per week. Compensation is $110 per hour.

Responsibilities

• Fine-tune and deploy transformer-based natural language processing models to align clinical trial registry entries with corresponding published outcomes
• Implement rule-based logic to detect and classify discrepancies between registered and published outcomes
• Build and validate a pipeline that evaluates whether omitted subgroup analyses were statistically feasible
• Integrate computational components into a unified evidence completeness framework
• Develop and document APIs supporting the prototype
• Containerize the final prototype using Docker
• Collaborate with the Principal Investigator, biostatistician, annotators, and software engineer
• Document technical decisions, model performance, validation results, and implementation requirements

Minimum qualifications

• Master’s or doctoral degree in computer science, data science, biomedical informatics, or a closely related field
• Demonstrated experience developing transformer-based NLP models, including BERT or domain-adapted variants
• Proficiency in Python, PyTorch, Hugging Face, and scikit-learn
• Experience developing APIs and containerizing applications with Docker
• Ability to work independently in a milestone-driven research and development environment
• Authorization to perform the work in the United States

Preferred qualifications

• Experience with biomedical text processing, clinical NLP, or health data
• Familiarity with ClinicalTrials.gov data structures and clinical trial publications
• Experience using AWS, GCP, or another cloud-based GPU training environment
• Experience supporting NSF SBIR, federally funded research, or regulated technology projects
• Experience developing reproducible and well-documented machine learning pipelines

University researchers, postdoctoral researchers, advanced graduate researchers who meet the minimum qualifications, and experienced independent professionals are encouraged to apply. Additional technical specifications will be shared with selected candidates.