GEN AI Engineer

tkxel; Lahore; On-site; PKR 1,200,000 - 1,800,000; Full time

Job summary

A tech company is looking for a skilled AI Developer in Lahore, Pakistan. The role focuses on developing and maintaining data pipelines for generative AI, requiring expertise in Python and AI/ML libraries, as well as cloud deployment experience. Candidates should have a relevant degree and 2-3 years of experience. Join an innovative team tackling complex AI challenges.

Qualifications

• 2-3 years of professional software development experience focused on AI/ML.
• Design and implement scalable backend services.
• Strong understanding of software engineering principles.

Responsibilities

• Design, build, and maintain robust data and model pipelines for generative AI.
• Develop, train, and fine-tune generative models.
• Architect and implement RAG systems for enhanced model accuracy.

Skills

Expert proficiency in Python

Experience with AI/ML libraries (e.g., PyTorch, TensorFlow)

Prompt Engineering for large language models (LLMs)

Familiarity with performance profiling and efficient model serving

Experience deploying models in a cloud environment (AWS, GCP, Azure)

Education

Bachelor's or Master's degree in Computer Science, AI, ML

Tools

Django

Flask

FastAPI

AWS SageMaker

Docker

Kubernetes

Job description

Key Responsibilities

• End-to-End Pipeline Development: Design, build, and maintain robust, scalable, and efficient data and model pipelines for generative AI applications, encompassing data ingestion, preprocessing, training, fine‑tuning, and inference.
• Model Training & Fine‑Tuning: Develop, train, and fine‑tune state‑of‑the‑art generative models (LLMs, LVMs, etc.) using advanced techniques such as Parameter‑Efficient Fine‑Tuning (PEFT/P‑Tuning/LoRA) and Reinforcement Learning from Human Feedback (RLHF).
• Retrieval‑Augmented Generation (RAG): Architect, implement, and optimize RAG systems by integrating vector databases (e.g., Pinecone, Milvus, Weaviate, pgvector) to enhance model accuracy and reduce hallucinations.
• Model Deployment: Package, containerize, and deploy models into production using modern MLOps and deployment tools like AWS SageMaker, Kubernetes, or similar platforms, ensuring reliability and scalability.
• Multi‑Modal & Agentic Development: Research and prototype multi‑modal (text, image, audio) AI solutions and develop agentic systems capable of planning and executing complex tasks.

Qualifications (Required)

• Education: Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Machine Learning, or a related technical field.
• Experience: 2‑3 years of professional software development experience with a strong focus on AI/ML.
• Design and implement scalable backend services using frameworks (Django, Flask, FastAPI).
• Programming: Expert proficiency in Python and strong experience with AI/ML libraries (e.g., PyTorch, TensorFlow, Hugging Face Transformers).
• Generative AI Fundamentals: Proven hands‑on experience in at least three of the following:
  • Prompt Engineering for large language models (LLMs).
  • Building applications using open‑source models from HuggingFace.
  • Implementing RAG architectures using vector databases.
  • Fine‑tuning models using PEFT methods (e.g., LoRA, Adapters).
  • Diffusion models, embeddings orchestration.
• Familiarity with performance profiling, efficient model serving, and hardware‑aware design (e.g., GPU utilization, quantization).
• Cloud & Deployment: Solid experience deploying and managing models in a cloud environment (AWS, GCP, or Azure). Direct experience with AWS SageMaker is a significant plus.
• Pipeline integration knowledge with web apps, for instance, Python and Ruby on Rails web applications.
• Software Engineering: Strong understanding of software engineering principles, design patterns, and writing production‑quality, maintainable code (version control, testing, debugging).

Qualifications (Preferred)

• Experience with RLHF and evaluating human preferences for model alignment.
• Practical knowledge of multi‑modal models (e.g., CLIP, FLAN‑T5, Vision Transformers).
• Experience building agentic workflows (e.g., using LangChain, LlamaIndex, or custom frameworks).
• Familiarity with containerization (Docker) and orchestration (Kubernetes).