Back to Portfolio

Internal AI Knowledge Assistant (Org-wide Chatbot Platform)

Software Engineer II

Chewy | 2023 – Present

Overview

Organization-wide chatbot platform leveraging MCP server architecture and multiple LLMs to help engineers and teams access internal knowledge and tooling.

The Problem

Engineers and teams needed faster access to internal documentation, workflows, and systems without manually digging through multiple tools and repos.

The Solution

Engineered an internal chatbot platform using MCP server architecture and integrated Anthropic Claude, OpenAI models, and internal systems to provide contextual answers and actions.

Impact

Improved cross-team productivity and knowledge sharing by centralizing access to documentation and workflows via conversational interfaces.

Key Responsibilities

  • Designed the architecture of the chatbot platform and MCP-based integrations.
  • Integrated multiple LLM providers and internal APIs while handling auth and rate limiting.
  • Worked with multiple teams to onboard domain-specific tools and knowledge sources.
  • Defined prompts, safety constraints, and evaluation strategies for internal use cases.

Technologies

TypeScriptNode.jsMCPAnthropic ClaudeOpenAI APIs