Gofer is an internal SAAS tool that we built to transform customer support KPIs and improve the effectiveness of support teams.

Support team KPIs are mainly focused on quantity over quality.

Assessing the performance of a support team is time-consuming and hard to do at scale.

Coaching support teams is costly, infrequent, and has decreasing returns.

An AI-driven tool that integrates with most ticketing tools.

Gofer evaluates in real-time the sentiment of each customer message and calculates the sentiment improvement rate per thread.

Gofer consolidates the sentiment data and provides insights into the company's effectiveness at addressing customer issues. This data is also broken down by agent to evaluate their effectiveness.

Gofer categories automatically every incoming message to offer insight into what aspects of the business require attention.

Gofer automatically summarizes large message threads so that managers can quickly understand what they are about.

Gofer also offers live coaching to support agents and gives them feedback on their responses to ensure that they are up-to standards.

For the Front end, we used SvelteKit, Tailwind, and Shadcn-svelte.

The Back End was built on Node.js with PostGres serving as the main DataBase.

We used a mixture of GCP's AI tools, Azure, LangChain, and Maritan LLM router for all of the AI functionality.

The entire solutions was deployed on Vercel, with the Back-End being deployed as serverless functions.