Machine Learning Engineer 2 - NLP

As an MLE-2 in NLP, you will independently own ML modules for applications such as search, summarization, dialogue systems, and text classification. You'll fine-tune LLMs, build evaluation pipelines, and contribute to applied research exploration.

Requirements:

• Experience: 3 Years to 7 years.

Programming and Frameworks:

• Strong proficiency in Python.
• Experience managing model training pipelines.
• Builds reusable components (tokenization, caching, distributed training).

NLP Expertise:

• Fine-tuning BERT, GPT, T5 and LLaMA for custom tasks.
• Strong understanding of seq2seq, attention, masking, and prompt engineering.
• Hands-on with NER, QA, summarization, and retrieval-augmented generation.

Evaluation and Experimentation:

• Familiarity with BLEU, ROUGE, perplexity, and diversity scores.
• Designed and executed A/B tests, qualitative user studies, and blind evaluations.
• Built interpretable error analysis dashboards.

Deployment and Ops:

• Deployed NLP APIs (FastAPI, Flask, TorchServe).
• Knowledge of ONNX, quantization, and LoRA (low-rank adaptation).
• Set up experiment tracking and hyperparameter optimization.

Soft Skills Expectations:

• Owns the roadmap for a submodule (e. g., summarization engine, feedback scorer).
• Actively mentors MLE-1s/interns.
• Collaborates with data, product, and infra teams to scope, deliver, and ensure adoption of ML features.