Senior LLM Engineer

Roles and Responsibilities:

• Model Expertise : Work with transformer models such as GPT , BERT , T5 , RoBERTa , and others for a variety of NLP tasks, including text generation, summarization, classification, and translation.
• Model Fine-Tuning : Fine-tune pre-trained models on domain-specific datasets to improve performance for specific applications such as summarization, text generation, and question answering.
• Prompt Engineering : Craft clear, concise, and contextually relevant prompts to guide transformer-based models towards generating desired outputs for specific tasks. Iterate on prompts to optimize model performance.
• Instruction-Based Prompting : Implement instruction-based prompting to guide the model toward achieving specific goals, ensuring that the outputs are contextually accurate and aligned with task objectives.
• Zero-shot, Few-shot, Many-shot Learning : Utilize zero-shot , few-shot , and many-shot learning techniques to improve model performance without the need for full retraining.
• Chain-of-Thought (CoT) Prompting : Implement Chain-of-Thought (CoT) prompting to guide models through complex reasoning tasks, ensuring that the outputs are logically structured and provide step-by-step explanations.
• Model Evaluation : Use evaluation metrics such as BLEU , ROUGE , and other relevant metrics to assess and improve the performance of models for various NLP tasks.
• Model Deployment : Support the deployment of trained models into production environments and integrate them into existing systems for real-time applications.
• Bias Awareness : Be aware of and mitigate issues related to bias , hallucinations , and knowledge cutoffs in LLMs, ensuring high-quality and reliable outputs.
• Collaboration : Collaborate with cross-functional teams including engineers, data scientists, and product managers to deliver efficient and scalable NLP solutions.

Must Have Skill

• Overall 7 years with at least 5+ years of experience working with transformer-based models and NLP tasks , with a focus on text generation , summarization , question answering , classification , and similar tasks.
• Expertise in transformer models like GPT (Generative Pre-trained Transformer) , BERT (Bidirectional Encoder Representations from Transformers) , T5 (Text-to-Text Transfer Transformer) , RoBERTa , and similar models.
• Familiarity with model architectures, attention mechanisms, and self-attention layers that enable LLMs to generate human-like text.
• Experience in fine-tuning pre-trained models on domain-specific datasets for tasks such as text generation , summarization , question answering , classification , and translation .
• Familiarity with concepts like attention mechanisms , context windows , tokenization , and embedding layers .
• Awareness of biases , hallucinations , and knowledge cutoffs that can affect LLM performance and output quality.
• Expertise in crafting clear, concise, and contextually relevant prompts to guide LLMs towards generating desired outputs.
• Experience in instruction-based prompting
• Use of zero-shot , few-shot , and many-shot learning techniques for maximizing model performance without retraining.
• Experience in iterating on prompts to refine outputs, test model performance, and ensure consistent results.
• Crafting prompt templates for repetitive tasks, ensuring prompts are adaptable to different contexts and inputs.
• Expertise in chain-of-thought (CoT) prompting to guide LLMs through complex reasoning tasks by encouraging step-by-step breakdowns.
• Proficiency in Python and experience with NLP libraries (e.g., Hugging Face, SpaCy, NLTK).
• Experience with transformer-based models (e.g., GPT, BERT, T5) for text generation tasks.
• Experience in training, fine-tuning, and deploying machine learning models in an NLP context.
• Understanding of model evaluation metrics (e.g., BLEU, ROUGE)

Qualification:

• BE/B.Tech or Equivalent degree in Computer Science or related field.
• Excellent communication skills in English, both verbal and written