Oscar – AI agent development, AI API integration, AI telemetry

An AI-powered multi-agent recruitment system designed to automate candidate sourcing and evaluation based on specific job descriptions. The platform uses Claude SDK-driven sub-agents and reusable skills to analyze candidate data, streamline recruiter workflows, and significantly reduce the time required to identify high-quality candidates.

• Defined the architecture and implementation of the AI system;
• Implemented agent memory and context management;
• Deployed the AI agent to production;
• Defined agent architecture and context management.

An AI-powered chatbot built with LangChain and Retrieval-Augmented Generation (RAG) to provide accurate, context-aware answers from multiple databases and document sources. The system leveraged LLMs to process user queries, retrieve relevant information, and support scalable conversational interactions for high-volume usage.

• Defined the agent architecture;
• Refined prompt engineering strategies;
• Scaled the system in production to support thousands of users;
• Set up evaluation workflows in Langfuse.

An AI-driven insights generation platform designed to transform structured data into actionable business intelligence using LLMs. The system enabled users to analyze complex datasets more efficiently and support faster, data-informed decision-making through natural language interactions and automated insight generation.

• Defined AI use cases and solution approach;
• Set up deployment processes and infrastructure;
• Implemented frontend functionality.

A data pipeline ingestion system designed to collect and standardize events from multiple platforms and services for downstream AI processing. The solution enabled reliable data aggregation, schema consistency, and scalable event handling to support AI-driven analytics and automation workflows.

• Defined the event pipeline architecture;
• Defined schemas and validation rules.