I will create or optimize a rag chatbot for you

A
alzapata317
A
alzapata317
Alejandro Z

About this gig

Most "AI search" tools are a vector database bolted to a chatbot, and nobody ever checks if the answers are actually correct. I build production-grade RAG pipelines that combine semantic search and keyword search, rerank results with a cross encoder, and come with a real evaluation suite that measures accuracy in numbers (Recall, MRR, nDCG), not vibes.

That means when I tell you your system retrieves the right answer 92% of the time, I can prove it. And when I make a design decision, like choosing a smaller reranking model because it actually scored higher than a bigger one in testing, it's backed by data, not guesswork.

If you're building a customer support bot, an internal knowledge base, or searching over your own documents, I'll build you a system that's fast, accurate, and honest about its own performance. No hallucinated confidence. No black box.

Let's turn your documents into answers people can trust.

Get to know Alejandro Z

Alejandro Z

AI Infrastructure Backend Engineer

  • FromUnited States
  • Member sinceAug 2026
  • Languages

    English
I build retrieval systems and measure whether they actually work. Most of my recent work is RAG infrastructure in Python and Postgres, with a focus on evaluation. I came up through automation workflow platforms like n8n and am now working a layer deeper.

My Portfolio

Related tags