Nivid D.

Developing an automated AI interview platform with dynamic question generation. Implementing multimodal analysis of candidate responses using sentiment & content scoring. Building reporting pipelines delivering insights to hiring teams. Designing LLM workflows & Improved latency by roughly 15–20% through backend optimization.

Spearheaded development of an AI-powered Compliance Automation System using Retrieval-Augmented Generation (RAG) for precise and regulation-aligned document analysis.

Integrated OpenAI and Mistral models with Pinecone Vector Search to reduce hallucinations and improve semantic retrieval accuracy.

Designed and deployed a No-Code Compliance Flow Builder, enabling users to generate domain-specific regulatory workflows without technical knowledge.

Built a Real-Time Compliance Scoring Dashboard that applies AI-driven risk assessments and predictive analytics to ensure proactive governance.

Developed an Audit & Explainability Layer, offering transparent and legally defensible compliance insights using natural language justifications.

Independently created a suite of AI automation tools, including:

Custom Bots for content interaction and auto-responses

Autonomous AI Agents with CrewAI, capable of executing complex multi-step tasks

Custom webgenerator with templates

LLM Fine-Tuning Pipelines and RAG Implementations tailored for vertical-specific use cases

Developed an AI-driven YouTube analytics system using NLP, sentiment analysis and metadata processing Built scalable Python pipelines for video content analysis, topic modeling and engagement prediction Created dashboards and automated reports for video performance metrics Integrated YouTube API with custom ML model for content analysis & tracking