Generative AI - Engineer

We are seeking a talented and passionate Generative AI Engineer to join our team. In this role, you will spearhead the end-to-end lifecycle of foundational model development, employing state-of-the-art approaches in generative AI. Your role will entail focus on data- and compute-efficient learning approaches to tackle challenges characteristics to Indian context. Your responsibilities will include optimizing model training and inference pipelines, production ready model deployment, ensuring scalability through distributed systems, and fine-tuning models on domain adaptation.

You will collaborate with cross-functional teams to build robust AI stacks and integrate them seamlessly into production pipelines. Beyond research and experimentation, you will play a pivotal role in transforming state-of-the-art models into operational systems that deliver measurable impact. As a leader in this domain, you will collaborate with other technical team members and subject matter experts, document technical processes, and maintain clear, organized codebases to support innovation and reproducibility. This role is ideal for proactive individuals passionate about driving transformative advancements in generative AI and scaling solutions for real-world impact.

Responsibilities:

• Development and training of foundational models across modalities
• End to end lifecycle management of foundational model development from data curation to model deployment by collaborating with the core team members
• Conduct research to advance model accuracy and efficiency.
• Implement state-of-the-art AI techniques in Text/Speech and language processing.
• Collaborate with cross-functional teams to build robust AI stacks and integrate them seamlessly into production pipelines.
• Develop pipelines for debugging, CI/CD and observability of the development process.
• Demonstrated ability to lead projects and provide innovative solutions.
• Should document technical processes, model architectures, and experimental results., maintain clear and organized code repositories.

Minimum Requirements:

Proficiency in Python programming and familiarity with 3-4 from the list of tools specified below:

1. Foundational model libraries and frameworks (TensorFlow, PyTorch, HF Transformers, NeMo, etc)
2. Experience with distributed training (SLURM, Ray, Pytorch DDP, Deepspeed, NCCL, etc)
3. Inference servers (vLLM, Triton, Reva)
4. Version control systems and observability (Git, DVC, MLFlow, W&B, KubeFlow)
5. Data analysis and curation tools (Dask, Milvus, Apache Spark, Numpy)
6. Speech processing libraries (Librosa, TorchAudio, Fairseq 1/2, NeMo)
7. Text-to-Speech/Speech-to-Text tools (Whisper, Voicebox, VALL-E (X), HuBERT/Unitspeech)
8. Vision models and libraries (Detectorn 2, DINO, Torch Vision)
9. LLMOPs Tools, Dockers etc
10. Hands on experience with AI application libraries and frameworks (DSPy, Langgraph, langchain, llamaindex etc.)

Desirable Skills:

• Publication record or work under review in top tier core A/A* conferences related to AI and its subfields.
• Past experience in community mentoring and participation. Specifically, you have a reputation of more than 150 in stack overflow or similar forums.
• Your coding contribution is accepted in an open source repo related to AI with more than 50 contributors.