Aryan Dixit

Built a multi-agent autonomous content compliance platform using AutoGen + LangGraph that detects profanity, nudity, subtitle mismatch, audio abnormalities, black screens, and policy violations in streaming/video content. Improved compliance issue detection performance to 99.8% recall, significantly reducing manual review effort. Developed MCP-enabled AI agents capable of self-correction through feedback loops and automated decision workflows.

Software Engineer (Agentic AI) - Samsung - Delhi

(2025-02)

• Built a multi-agent autonomous content compliance platform using AutoGen + LangGraph that detects profanity, nudity, subtitle mismatch, audio abnormalities, black screens, and policy violations in streaming/video content.
• Improved compliance issue detection performance to 99.8% recall, significantly reducing manual review effort.
• Developed MCP-enabled AI agents capable of self-correction through feedback loops and automated decision workflows.
• Designed secure deployment workflows using AWS EC2, S3, Docker, and private infrastructure environments.
• Built AI-powered chatbot for Samsung TV Plus monitoring teams to analyse alarms and troubleshoot channel failures.
• Created real-time visualization dashboards using Plotly + React for monitoring stream health metrics.
• Worked on stream anomaly detection involving ad-break validation, black screen detection, freeze detection, and MSN mismatch detection.

Data Scientist Intern - NielsenIQ - Gurugram

(2024-07 - 2025-01)

• Built forecasting models for business prediction tasks using historical datasets.
• Automated perception calculation workflows and reduced manual reporting efforts.
• Perform EDA, data cleaning, dashboard reporting and business analytics.
• Created client-facing reports using Python, Power BI, SQL, and Excel.

BTech - Computer Science and Engineering with Specialization in Data Science - Sharda University (2021-09 - 2025-05)