ML Researcher, Foundational Models

Ovii's Interpretation of the Role

We are seeking an ML Researcher to lead open‑ended foundational model work, from hypothesis to large‑scale pre‑training experiments and production deployment. You will own the full research loop, collaborate with infrastructure and data teams, and publish impactful results.

Role Snapshot

• Lead open‑ended foundational model research
• Design and run large‑scale pre‑training ablations
• Translate findings into production training proposals
• Partner with infrastructure and data engineering
• Publish internally and at top‑tier venues
• Own end‑to‑end model training pipelines

Must-Have Requirements

• PhD in Machine Learning, Computer Science, or related field
• 3+ years post‑PhD research experience
• First‑author publications at top‑tier ML venues (NeurIPS, ICML, ICLR, ACL, EMNLP, COLM)
• Hands‑on pre‑training of transformer models 7B+ parameters
• Fluency in PyTorch and distributed training
• Strong intuition for experimental design and ablation
• post‑PhD research
• transformer pre‑training
• top‑tier publications
• PhD in Machine Learning, Computer Science, or closely related field
• PhD required

Nice-to-Have Signals

• Work on novel architectures (Mixture‑of‑Experts, state‑space, hybrid)
• Experience with multilingual or multimodal pre‑training
• Research contributions in post‑training (RLHF, RLVR, distillation, reasoning)
• Track record shipping research ideas to production capabilities
• Open‑source LLM ecosystem contributions (code, model releases, datasets)
• novel model architectures
• multilingual/multimodal pre‑training
• post‑training techniques

Work Setup

• Location: Bengaluru, India
• Work mode: ONSITE
• Remote scope: UNSPECIFIED
• Employment type: Full-Time

Not Specified in JD

• Visa sponsorship
• Salary range
• Remote eligibility
• Travel requirement
• Security clearance
• Coding test
• Portfolio

What You'll Likely Work On

• Drive research on model architecture, optimization, scaling behavior, training stability, and post‑training recipes
• Design and execute ablation experiments at scale that directly inform large‑run decisions
• Turn research outcomes into concrete proposals and shepherd them through production training runs
• Collaborate closely with infrastructure and data teams on research‑engineering boundary problems
• Read broadly, write internal reports, and publish externally when the work merits it

Good Fit If You Have

• Experience with novel model architectures such as mixture‑of‑experts or state‑space models
• Background in multilingual or multimodal pre‑training
• Track record taking a research idea from prototype to shipped capability
• Contributions to open‑source LLM ecosystems