Freelance Data Engineer

To those in my network, we're looking for some freelance help on a short-term project to convert an Excel dataset into an LLM-ready vector database as part of our exciting product development programme.

If you're interested or know someone you think would be, please get in touch!

Overview

We have a one-off project to turn a single Excel spreadsheet into a vectorised, LLM-queryable dataset. We'll provide the source file; you deliver a working vector store plus the script that produced it.

Scope of work

• Clean and prepare the data from our Excel file (deduplication, normalisation, consistent formatting).
• Chunk the records into sensible units for embedding.
• Generate vector embeddings using an appropriate model (advise us if you have a preference).
• Load the embeddings into a vector database of your recommendation (e.g. Chroma, Qdrant, pgvector, or Pinecone).
• Preserve source metadata so results can be filtered and traced back to original rows.
• Provide the reusable script/notebook so we can re-run it ourselves if the data changes.

Deliverables

• A populated, queryable vector store
• The commented pipeline script (Python)
• A short README covering how to run it, the model/DB chosen, and any costs (e.g. embedding API usage)
• A brief walkthrough or call to hand it over

Skills needed

• Python (pandas + embedding/vector libraries)
• Experience with embedding models and at least one vector database
• Solid data-cleaning instincts

If you’re interested in this opportunity please email recruiter contact, including why you would be ideal for this project and an indication of your dayrate.