Software Engineer

Description : Role Overview :We are seeking a highly skilled Software Engineer specializing in Large Language Models (LLMs) to design, develop, and deploy cutting-edge AI solutions leveraging state-of-the-art transformer architectures.- The ideal candidate will have strong expertise in deep learning, NLP, and model optimization, combined with software engineering best practices for building scalable AI systems in production.- Youll collaborate with data scientists, ML engineers, and product teams to build intelligent applications powered by advanced generative AI models such as GPT, LLaMA, Falcon, Mistral, Claude, or similar open-source and proprietary models.Key Responsibilities :- Design, train, fine-tune, and evaluate Large Language Models (LLMs) for specific use cases (e.g., summarization, code generation, chatbots, reasoning, and retrieval-augmented generation).- Experiment with transformer-based architectures (e.g., GPT, T5, BERT, LLaMA, Mistral).- Develop parameter-efficient fine-tuning (PEFT) strategies such as LoRA, QLoRA, adapters, or prompt-tuning.- Create and maintain high-quality datasets for pretraining, fine-tuning, and evaluation.- Optimize model inference using techniques like quantization, distillation, and tensor parallelism for real-time or edge deployment.- Integrate LLMs into production environments using frameworks like Hugging Face Transformers, PyTorch Lightning, or DeepSpeed.- Implement scalable model serving solutions using FastAPI, Ray Serve, Triton Inference Server, or similar frameworks.- Build and maintain APIs or SDKs that expose LLM capabilities to other teams and products.- Evaluate and experiment with open-source and proprietary foundation models.- Keep up with the latest trends in Generative AI, NLP, and Transformer models.- Perform benchmarking, ablation studies, and A/B testing to measure performance, cost, and quality improvements.- Collaborate with ML Ops and DevOps teams to design CI/CD pipelines for model training and deployment.- Manage and optimize GPU/TPU clusters for distributed training and inference.- Implement robust monitoring, logging, and alerting for deployed AI systems.- Ensure software follows clean code principles, version control, and proper documentation.- Partner with product managers, data scientists, and UX teams to identify and translate business problems into AI-driven solutions.- Contribute to internal research initiatives and help shape the companys AI strategy.- Mentor junior engineers in AI model development, coding standards, and best practices.Required Technical Skills : Core Expertise : - Strong proficiency in Python and deep learning frameworks (PyTorch, TensorFlow, JAX).- Hands-on experience with transformer architectures and LLM fine-tuning.- Deep understanding of tokenization, attention mechanisms, embeddings, and sequence modeling.- Experience with Hugging Face Transformers, LangChain, LlamaIndex, or OpenAI API.- Experience deploying models using Docker, Kubernetes, or cloud ML services (AWS Sagemaker, GCP Vertex AI, Azure ML, OCI Data Science).- Familiarity with model optimization (quantization, pruning, distillation).- Knowledge of retrieval-augmented generation (RAG) pipelines, vector databases (FAISS, Pinecone, Weaviate, Chroma).Additional Skills (Good to Have) : - Experience with multi-modal models (text + image, text + code).- Familiarity with MLOps tools like MLflow, Kubeflow, or Weights & Biases (W&B).- Understanding of Responsible AI practicesbias mitigation, data privacy, and model explainability.- Experience contributing to open-source AI projects (ref:hirist.tech)