At Calsoft, we’re constantly innovating to give our business development team the edge they need. Today, I’m excited to share how we’ve leveraged cutting-edge AI technologies to create CalPitch, our new sales outreach tool that’s transforming the way we connect with prospects.
CalPitch is built using LlamaIndex and OpenAI’s GPT models. These technologies work in tandem to create personalized, context-aware email pitches that resonate with our potential clients. Here’s how we’ve integrated these powerful tools:
1. Document Indexing: We use LlamaIndex to create and manage a vector store index of our company documents, case studies, and service descriptions. This allows us to quickly retrieve relevant information for each prospect.
2. Contextual Querying: The `chat_engine` from LlamaIndex enables us to perform contextual queries against our knowledge base, ensuring that every email pitch is grounded in Calsoft’s expertise and offerings.
3. Efficient Storage: LlamaIndex’s integration with FAISS (Facebook AI Similarity Search) allows us to store and retrieve high-dimensional vectors efficiently, making our system scalable and fast.
1. GPT-4o for Email Generation: We leverage OpenAI’s GPT-4o model to generate highly personalized email content. The model takes into account the prospect’s information, company details, and our custom instructions to craft compelling pitches.
2. Embeddings: OpenAI’s embedding model is used to convert text into high-dimensional vectors, enabling semantic search and similarity comparisons within our knowledge base.
1. Data Ingestion: The system starts by ingesting prospect data, either from a CSV file or a LinkedIn URL.
2. Enrichment: We use the Proxycurl API to enrich the prospect’s data with detailed information about their role and company.
3. Context Retrieval: LlamaIndex’s chat engine queries our knowledge base to find relevant information about Calsoft’s services that align with the prospect’s needs.
4. Email Generation: The enriched prospect data and retrieved context are fed into OpenAI’s GPT-4 model, which generates a personalized email pitch.
5. Review and Sending: Our BD team can review, edit if necessary, and send the generated emails to prospects.
1. Time Efficiency: CalPitch automates the time-consuming process of researching prospects and crafting personalized emails, allowing our BD team to focus on high-value interactions.
2. Consistency: Every email maintains a consistent tone and quality, aligned with Calsoft’s brand voice.
3. Personalization at Scale: We can now create highly personalized pitches for many prospects without sacrificing quality.
4. Data-Driven Approach: By leveraging our entire knowledge base, we ensure that every pitch is backed by our most relevant case studies and expertise.
5. Continuous Improvement: The system learns from feedback, allowing us to refine our approach over time.
We’re seen over 10k+ emails generated since we’ve started building CalPitch a few of months ago. BD team members who use CalPitch have better conversation rates than BD team members who don’t, which increase productivity of our teams massively.
CalPitch is just the beginning. As we continue to refine and expand this tool, we’re excited about the possibilities it opens for our business development efforts. By combining the power of LlamaIndex and OpenAI with our team’s expertise, we’re not just keeping up with the future of sales outreach – we’re defining it.
Stay tuned for more innovations from Calsoft as we continue to push the boundaries of what’s possible in IT services and sales technology!
Check out our services to learn how we can efficiently push AI solutions to production and how we specialize in Gen AI, Analytics & ML, and Data to support enterprises and ISVs.
Llamaindex has made the process of going from idea to production extremely straightforward. The flexibility of deciding which type of RAG pipeline to allows us much more control over how we can shape my applications and workflows, and combined with their extension integrations, makes our code work with other tools much more efficient.
Rachitt Shah, AI Engineer, Calsoft
