Chatbot Text Generation NLP

I’m building a conversational assistant and need an AI engineer who can take charge of the full Natural Language Processing pipeline for high-quality, context-aware chatbot responses. The core of the project is text generation: designing, training, and fine-tuning a model that answers user prompts fluidly, stays on topic, and preserves an appropriate tone of voice.

You’ll work with my existing dialogue data (plus any open-source corpora you recommend) to create a model that can:

• Understand multi-turn context and user intent
• Respond naturally in English without hallucinating facts
• Respect soft constraints I’ll provide on length, formality, and persona

Typical tools in this space—Python, PyTorch or TensorFlow, Hugging Face Transformers, and popular evaluation libraries—fit well here, but I’m flexible if you have a stronger stack.

Deliverables
1. Pre-processed, reproducible dataset and accompanying scripts
2. Fine-tuned model checkpoints with clear versioning
3. Inference wrapper (REST API or lightweight microservice) that I can drop into my backend
4. Short walkthrough document and recorded demo showing the system answering sample queries

I’ll validate the work with BLEU / ROUGE metrics plus live tests against real chat logs; final acceptance is a model that meets the agreed quality benchmarks and runs on a single GPU. If this sounds straightforward to you, let’s talk timeline and milestones so we can get started right away.