RAG developer | Gen AI startup

Join serenitygpt.com as a Gen AI developer where you'll have the opportunity to take ownership of cutting-edge projects that leverage Generative AI to transform search technology and automate customer support. Work with a dynamic and ambitious team committed to doing meaningful work and solving real-world problems for global tech and BPO companies. This role challenges you to push your limits, collaborate with talented peers, and grow both personally and professionally as we revolutionize the way search and support systems function. Embrace the chance to be part of an early-stage startup that has just proven traction.

About Us We are a Gen AI startup that actually has customers. We fix a problem that resonates with everyone: it's hard to find information and almost every search box is broken. We make search boxes understand human language so users can ask questions in human language and get concise verifiable answers with links to sources. We target two use cases:

search over documentation, knowledge base articles, and technical forums (we work with global tech companies)
automating ticket-based customer support: a ticket comes in, we find an answer based on KBs, prior tickets, support channels in Slack, etc (we work with global BPO companies offering support)

Here is a 20-minute video describing what we do: https://www.youtube.com/watch?v=GcdhMoJWikg

Key Responsibilities Create Retrieval Augmented Generation (RAG) pipelines that work best for the use case at hand. Degrees of freedom:

chunking strategy (how to split documents before indexing)
embedding model
metadata to include in each chunk
LLM
LLM prompt, including Chain of Thought and multi-stage reasoning

The key to making decisions about the above is having a robust evaluation framework with good eval sets

develop an evaluation framework for search (i.e., retrieval) and for search+answer (i.e., for end-to-end retrieval and generation)
create and curate eval sets for both search and LLM
see our own blog post for some basic ideas on how to create an eval set for embedding models https://serenitygpt.com/blog/how-to-choose-the-embedding-model-for-rag

Why us - We are ambitious, professional, and human (as opposed to corporate)

We are on the cutting edge of the most exciting (both in terms of hype and technology) breakthrough of the last couple of decades - Gen AI
We work on concrete use cases that are in production with customers. This gives us the ability to evaluate our system on real use cases
We have 100k in credits from Google Cloud to use on compute and any of the AI models available through Google Cloud
We have 25k in Azure credits to use on compute and any of the AI models available through Azure including gpt-4o
We want you to succeed and we'd love to grow together with you

Requirements Industry Experience:

At least 3 years of industry/startup experience ideally in a fast-paced environment
Significant experience with RAG / embedding models / information retrieval

Stack