Job Title:
Data Scientist – SDE-3 (Generative AI / LLMs)
Location:
Noida / Bangalore
Qualification & Eligibility
Bachelor’s or Master’s degree in Computer Science, Statistics, Engineering, Applied Mathematics, or a related quantitative field.
Strong theoretical and practical grounding in machine learning, deep learning, and software engineering.
Experience
5–8 years of experience
in Data Science / ML Engineering with a strong focus on deep learning and LLM-driven systems.
Demonstrated experience
leading complex AI projects
end-to-end, including architecture, model design, and production deployment.
Startup experience preferred; EdTech or other high-scale consumer product experience is a strong advantage.
Roles & Responsibilities
Architect, design, and deploy
large-scale solutions leveraging LLMs for production use cases.
Own the full lifecycle: research, experimentation, model development, evaluation, fine-tuning, optimization, deployment, and monitoring.
Lead advanced research in
Generative AI, LLMs, Fine-Tuning (LoRA/QLoRA), Reinforcement Learning (RLHF/RLAIF), RAG, and agentic systems
.
Build scalable
frameworks, libraries, and reusable pipelines
to accelerate GenAI development across the company.
Take part in high-impact technical design discussions, conduct structured code reviews, and drive engineering excellence.
Collaborate cross-functionally with product, engineering, and research teams to integrate AI solutions into
core platforms at scale
.
Demonstrate strong analytical thinking and ability to break down ambiguous problems into actionable engineering steps.
Mentor junior engineers and contribute to building a high-performance AI engineering culture.
Required Skills:
Core Technical Expertise
Strong command over
Deep Learning architectures
– Transformers, CNNs, RNNs, Attention mechanisms.
Expert-level proficiency in
LLMs & Generative AI
, including:
Fine-Tuning techniques:
LoRA, QLoRA, PEFT, instruction tuning
Reinforcement Learning:
RLHF, RLAIF, preference optimization
Prompt Engineering:
RAG, CoT, ToT, ReAct, agent frameworks
LangChain, LlamaIndex, Vector Databases
Ecosystem & Tooling
Hands-on experience with
Azure OpenAI, Google Vertex AI, AWS Bedrock
across multimodal workloads (text/audio/image/video).
Strong understanding of
Open-Source LLMs
(Llama, Mistral, Gemma, Mixtral) and Hugging Face pipelines.
Experience with
PyTorch / TensorFlow
, Transformers library, model quantization and optimization.
Proficiency in
Python
and backend frameworks such as FastAPI, Django, Flask.
Practical experience with
Docker, Git, CI/CD pipelines
, and cloud infrastructure (Azure/AWS/GCP).
Company Description
PhysicsWallah (PW) is India’s leading edtech platform, founded in 2016 by Alakh Pandey with a mission to make high-quality education accessible and affordable. Starting as a YouTube channel, PW has grown into a
multi-billion-dollar unicorn
serving millions of learners through online courses, offline hybrid centres, and test-prep programs for JEE, NEET, UPSC, state boards, and more.
With
15M+ app downloads
,
2M+ daily learners
, and
120+ offline centres
, PW continues to transform learning through technology, innovation, and affordability. Recognised as a
Great Place to Work
, PW is rapidly expanding globally and is investing heavily in
AI-driven educational innovation
.