Staff AI Engineer

PW (PhysicsWallah); Dadri; On-site; INR 4,000,000 - 8,000,000; Full time

Job summary

PhysicsWallah invites an AI Staff Data Scientist to lead Generative AI and LLM initiatives. You will design production-grade AI systems, optimize inference latency, and build scalable RAG pipelines for knowledge-intensive tasks.

You will mentor engineers, collaborate across Product, Engineering and Data Science teams, and stay ahead of advances in AI model serving, vector databases, and agentic architectures.

Qualifications

• Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Data Science, or a related field.
• 7–12+ years of experience in Machine Learning, Data Science, AI Engineering, or related domains.
• Proven experience building and deploying AI solutions in production environments; strong knowledge of distributed systems, scalable architectures and cloud platforms.

Responsibilities

• Design, develop, and deploy production-grade Generative AI and LLM-powered applications.
• Architect and implement scalable RAG pipelines for knowledge-intensive AI systems.
• Fine-tune foundation models using LoRA, QLoRA, PEFT, and RLHF.
• Optimize LLM inference for latency, throughput, scalability, and infrastructure cost.
• Build Agentic AI and multi-agent systems capable of autonomous reasoning and task execution.
• Design scalable vector search architectures using modern vector databases.
• Develop robust evaluation pipelines, including LLM-as-a-Judge methodologies for automated quality assessment.
• Build and maintain end-to-end ML workflows, experimentation pipelines, and model observability.
• Collaborate with Product Managers, Engineering teams, and Data Scientists to translate business problems into AI-powered solutions.
• Drive architectural decisions, establish AI best practices, and mentor engineers across teams.
• Stay up to date with the latest advancements in Generative AI, LLM optimization, and model serving technologies.

Skills

vLLM

Text Gen Inference

SGLang

RAG

LoRA

QLoRA

PEFT

RLHF

LLM-as-a-Judge

Quantization

Distillation

Education

Bachelor/Master in CS/AI/DS

Tools

LangChain

LangGraph

Pinecone

Weaviate

Hugging Face

PyTorch

Flash Attention

LLM Serving

Job description

PhysicsWallah (PW) is one of India's leading education technology companies, empowering millions of learners through innovative digital learning solutions. We are building next-generation AI platforms to transform education through personalized learning, intelligent tutoring, content generation, and AI-powered learning experiences.

As an AI Staff Data Scientist, you will play a key technical leadership role in shaping our Generative AI and Large Language Model (LLM) initiatives. You will work closely with Product, Engineering, and AI teams to build scalable, production-grade AI systems that power experiences for millions of students.

If you're passionate about solving complex AI challenges and building state-of-the-art LLM applications at scale, we'd love to hear from you.

Responsibilities

• Design, develop, and deploy production-grade Generative AI and LLM-powered applications.
• Architect and implement scalable Retrieval-Augmented Generation (RAG) pipelines for knowledge-intensive AI systems.
• Fine-tune foundation models using techniques such as LoRA, QLoRA, PEFT, and RLHF.
• Optimize LLM inference for latency, throughput, scalability, and infrastructure cost.
• Build intelligent Agentic AI and multi-agent systems capable of autonomous reasoning and task execution.
• Design scalable vector search architectures using modern vector databases.
• Develop robust evaluation pipelines, including LLM-as-a-Judge methodologies for automated quality assessment.
• Build and maintain end-to-end ML workflows, experimentation pipelines, and model observability.
• Collaborate with Product Managers, Engineering teams, and Data Scientists to translate business problems into AI-powered solutions.
• Drive architectural decisions, establish AI best practices, and mentor engineers across teams.
• Stay up to date with the latest advancements in Generative AI, LLM optimization, and model serving technologies.

Skills & Qualifications

Required Skills

• Strong experience designing and deploying production-grade LLM applications.
• Hands-on expertise with:
• vLLM or Text Generation Inference (TGI) or SGLang
• Retrieval-Augmented Generation (RAG)
• LoRA, QLoRA, or PEFT
• Reinforcement Learning from Human Feedback (RLHF)
• Quantization and/or Distillation
• LLM-as-a-Judge evaluation frameworks

AI Frameworks

• LangChain
• LangGraph

Vector Databases

• Pinecone
• Weaviate
• Hugging Face Transformers
• PyTorch
• Flash Attention

Agentic AI

• AI Agents
• Multi-Agent Systems

Experience with

• MLflow
• Langfuse
• Weights & Biases (W&B)

Performance Optimization

• LLM latency optimization
• LLM inference optimization
• Cost optimization for production AI systems

Qualifications

• Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Machine Learning, Data Science, or a related field.
• 7–12+ years of experience in Machine Learning, Data Science, AI Engineering, or related domains.
• Proven experience building and deploying AI solutions in production environments.Strong understanding of distributed systems, scalable architectures, and cloud platforms (AWS, Azure, or GCP).
• Excellent problem-solving, communication, and stakeholder management skills.
• Experience mentoring engineers and driving technical initiatives is highly preferred.

Why Join PhysicsWallah?

• Build AI products that impact millions of learners across India.
• Work on cutting-edge technologies in Generative AI, LLMs, Agentic AI, and RAG.
• Solve challenging real-world problems with large-scale AI systems.
• Collaborate with high-performing Product, Engineering, and AI teams.
• Enjoy a fast-paced, ownership-driven culture where innovation is encouraged.
• Lead strategic AI initiatives and influence the future of AI-powered education.
• Accelerate your career by working on some of the most impactful AI use cases in the EdTech ecosystem.