About BuildingBlocks:

Regardless of whether accomplishing astounding customer work or aiding assemble our organization from the inside, our different gifted organization shares a solitary mission: to fabricate lovely encounters that decidedly affect individuals lives and the organizations we serve.

Our way of life is more reason than work, mixing the quest for certified human association with a fastidious way to deal with the things we make. All with an idealistic soul. Were an organization that is adequately youthful to be fun and defiant - yet intense and driven enough to immensely affect the business.

Role Overview:

You will own the data pipeline powering our LLM training and fine-tuning. This includes ingestion, cleaning, deduplication, and building high-quality datasets from structured/unstructured sources.

Responsibilities:

- Design ETL pipelines for text, PDFs, and structured data.

- Implement data deduplication, filtering (toxicity, PII), and normalization.

- Train and manage tokenizers (SentencePiece/BPE).

- Build datasets for supervised fine-tuning and evaluation.

- Work closely with domain experts to generate instruction/response pairs.

Requirements:

- Strong in Python, SQL, and data wrangling frameworks (Pandas, Spark).

- Experience with large text datasets, cleaning, preprocessing.

- Familiarity with NLP-specific preprocessing (chunking, embeddings).

- Knowledge of cloud data storage (S3/GCS/Blob).

- Bonus: Prior experience in AI/ML pipelines.