AI Developer (Karachi Onsite)

We are looking for a talented AI/ML Engineer with 2 to 3 years of practical experience to join our team. In this role, you will be responsible for designing, fine-tuning, and deploying state-of-the-art Generative AI solutions. Your focus will be on optimizing Large Language Model (LLM) architectures for efficient inference, applying advanced quantization techniques, and developing robust multi-modal and multi-lingual AI pipelines. This position offers the opportunity to work on cutting-edge AI technologies and contribute to scalable production systems.

Key Responsibilities
- Architect, develop, and train machine learning models, primarily Large Language Models (LLMs).
- Perform end-to-end fine-tuning workflows such as LoRA, QLoRA, and DPO to customize pre-trained models for specific domains.
- Implement model compression and quantization techniques (including bitsandbytes, AWQ, GGUF) to optimize models for production deployment.
- Build and integrate multi-modal systems that handle text, image, and audio data, while supporting multi-lingual capabilities for global applications.
- Deploy models into production environments using tools like Triton Inference Server and vLLM, ensuring scalable and reliable infrastructure.
- Analyze and troubleshoot Transformer-based architectures, including Attention mechanisms, Mixture of Experts (MoE), and KV-caching, to improve model performance.

Required Qualifications
- 2 to 3 years of hands-on experience in AI/ML engineering, with a focus on Generative AI and Large Language Models.
- Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, or a related field.
- Proficiency in PyTorch or TensorFlow frameworks.
- Strong familiarity with the Hugging Face ecosystem, including Transformers, PEFT, Accelerate, and TRL.
- Experience with containerization technologies such as Docker and Kubernetes, and model serving tools like NVIDIA Triton, vLLM, or FastAPI.
- Solid understanding of GPU-accelerated computing, including CUDA and NVLink configurations.
- Practical knowledge of multi-modal architectures such as CLIP and LLaVA.
- Experience with multi-lingual tokenization and strategies for adapting models across languages.
- Strong problem-solving skills, with the ability to balance trade-offs between model size, inference speed, and accuracy.

Preferred Qualifications and Benefits
- Experience developing agentic AI workflows using tools like LangGraph or CrewAI.
- Prior work on Retrieval-Augmented Generation (RAG) systems, including designing production-grade pipelines with vector databases such as Milvus, Qdrant, or Pinecone.
- Contributions to open-source AI projects or a portfolio showcasing research and development in AI models and frameworks.
- Competitive monthly salary range of Rs150,000 to Rs250,000.
- This is an in-person role, offering the chance to collaborate closely with a dynamic team on innovative AI projects.

This position is ideal for professionals passionate about pushing the boundaries of AI technology and eager to work in a fast-paced, innovative environment.