For a data and analytics consulting agency, I developed an AI Agent to support sales, project delivery, and general document generation activities. It included a Knowledge Base powered by Pinecone and OpenAI embeddings, continuously updated in real time using an n8n workflow that monitors a Google Drive repository. A multi-step agent built with Langflow and GPT-4o and integrated with Fireflies.ai and Google Docs enabled automated creation of consulting reports and customised sales materials. Delivered via self hosted LangFlow with Docker on GCP exposing a Chatbot for end users.

Built for a Data and Analytics consulting Agency to support their sales, marketing, and document generation activities

A Knowledge Base containing a large repository of past documents and sales and marketing materials was built using Pinecone and OpenAI embedding models

A n8n workflow was implemented to continuously update the Knowledge Base in real time by monitoring a google drive containing all the documents, and dynamically loading new documents into Pinecone and updating the embeddings

A multi-step agent was built using Langflow and GPT-4o, with pre-built flows for specific types of content generation, i.e. sales presentation, consulting report, marketing 1-pager

Multiple systems were integrated for different workflows, i.e. Fireflies.ai to generate consulting reports from meeting transcripts, Google Docs for generating customised client sales materials based on previous documents 

The client’s team interacted via a ChatBot, leveraging Langflow’s inbuilt capabilities and hosted via Docker on GCP

Langflow, OpenAI, GPT-4o, n8n, Pinecone, Fireflies.ai, Meeting Transcription, Google Docs, GCP, Python