7 AI-POWERED BACKGROUND CHECK / ANONYMIZER USING SNOWFLAKE CORTEX AGENT PFE

Project Overview

• Develop an AI-powered agent leveraging Snowflake Cortex to automate background verification workflows.
• Aggregate and correlate data from multiple trusted sources while ensuring data privacy through anonymization and pseudonymization techniques.
• Team size: 2 interns. Estimated duration: 4-6 months.

Objectives & Responsibilities

• Design and implement a Snowflake Cortex agent that can query, normalize and enrich data from external and internal trusted sources.
• Implement anonymization and pseudonymization mechanisms to protect personal data while preserving verification utility.
• Integrate OpenAI models for intelligent parsing, entity resolution, and risk scoring where appropriate.
• Build API endpoints and backend services using .NET to orchestrate data flows between Snowflake, AI models and front-end components.
• Collaborate with a front-end developer to expose verification results and anonymization controls in an Angular UI.

Technical Stack & Architecture

• Core technologies: Snowflake Cortex, OpenAI, .NET (backend), Angular (frontend), Azure cloud services.
• Key components: data ingestion pipelines, Cortex agent logic, anonymization/pseudonymization module, AI-driven entity resolution and scoring, secure storage and logging on Azure.

Requirements & Skills Sought

• Familiarity with Snowflake (preferably Cortex) and data engineering concepts for secure data aggregation.
• Experience or strong interest in applied ML / LLM integration (OpenAI or similar) for text understanding and entity extraction.
• Proficiency in .NET for backend APIs and Angular for front-end integration is highly valued.
• Knowledge of privacy-preserving techniques (k-anonymity, pseudonymization, hashing, tokenization) and compliance considerations.

Deliverables & Evaluation

• A working Snowflake Cortex agent prototype that performs multi-source background checks with privacy-preserving outputs.
• End-to-end demo including .NET APIs and an Angular interface showcasing anonymized verification results and audit logs.
• Documentation: design decisions, data flow diagrams, privacy measures implemented, and test/evaluation results.

How to Apply

• To apply, send your application referencing this specific topic to the recruitment email: recrutementunisie@finlogik.com.
• You can also apply via the online posting: https://secure.collage.co/jobs/finsotech/57794