Software Engineer - Python (AI)

Role Overview: You will be a hands-on AI engineer responsible for designing, prototyping, and delivering generative AI capabilities. Your role will involve practical research, building POCs, fine-tuning open models, contributing to multimodal experiments, and assisting in taking solutions towards production. You will have the opportunity to rapidly learn modern and classical ML techniques. Key Responsibilities: - Build and evaluate prototypes / POCs for generative AI features and ideas. - Fine-tune and adapt open-source LLMs and smaller generative models for targeted use cases. - Collaborate on multimodal experiments involving text, image, and audio, and implement training/evaluation pipelines. - Implement data preprocessing, augmentation, and basic feature engineering for model inputs. - Run experiments by designing evaluation metrics, performing ablations, logging results, and iterating on models. - Optimize inference and memory footprint for models through quantization, batching, and basic distillation. - Contribute to model training pipelines, scripting, and reproducible experiments. - Work with cross-functional teams (product, infra, MLOps) to prepare prototypes for deployment. - Write clear documentation, present technical results to the team, participate in code reviews, and share knowledge. - Continuously learn by reading papers, trying new tools, and bringing fresh ideas into projects. Qualification Required: Mandatory Technical Skills: - Strong Python programming skills and familiarity with ML tooling such as numpy, pandas, and scikit-learn. - Hands-on experience (2+ years) with PyTorch and/or TensorFlow for model development and fine-tuning. - Solid understanding of classical ML & DL concepts including supervised/unsupervised learning, optimization, CNNs, RNNs/LSTMs, and Transformers. - Good knowledge of algorithms & data structures, numerical stability, and computational complexity. - Practical experience in fine-tuning open models like Hugging Face Transformers, LLaMA family, BLOOM, Mistral, or similar. - Familiarity with PEFT approaches (LoRA, adapters, QLoRA basics) and simple efficiency techniques like mixed precision and model quantization. - Comfortable with running experiments, logging using tools like Weights & Biases and MLflow, and reproducing results. - Exposure to at least one cloud ML environment (GCP Vertex AI, AWS SageMaker, or Azure AI) for training or deployment tasks. - Good communication skills for documenting experiments and collaborating with product/infra teams. Highly Desirable / Preferred Skills: - Experience with multimodal training pipelines or cross-modal loss functions. - Familiarity with MLOps concepts such as model packaging, CI/CD for models, and basic monitoring. - Experience with tools like DeepSpeed, Accelerate, Ray, or similar distributed/efficiency libraries. - Knowledge of LangGraph / Autogen / CrewAI or interest in agentic systems. - Experience with BigQuery / Synapse or data warehousing for analytics. - Publications, open-source contributions, or sample projects demonstrating model work (GitHub, Colabs, demos). - Awareness of AI safety and responsible-AI best practices. (Note: Additional details about the company were not provided in the job description.),