AI/ML Engineer – Foundation Models & Infra
Company Overview
It is partnering with an early-stage startup to hire a Founding AI/ML Engineer who will work closely with the founding team to build and scale cutting-edge AI products from the ground up.
Job Summary
The AI/ML Engineer will be responsible for training and fine-tuning machine learning models as well as establishing the infrastructure to ensure reliable execution. This role is pivotal in enhancing the model layer assessment engine, which involves optimizing models for risk classification and adversarial evaluation, maintaining fast and cost-effective training/inference infrastructure, and collaborating closely with the company's founders.
Responsibilities
- Fine-tune open weight models for risk assessment, adversarial/jailbreak detection, and evaluation of AI agents.
- Design and own the training infrastructure, including data pipelines, distributed training setup, experiment tracking, reproducibility, and executing the models.
- Optimize for quantization, distillation, and inference tuning, balancing latency, throughput, and cost.
- Stand up and manage GPU/compute infrastructure (cloud and/or self-hosted), focusing on cost modeling and capacity planning.
- Design evaluation harnesses to measure model quality, robustness, and drift per the company’s Risk Framework.
- Deploy, serve, and monitor models in production, managing throttling, scaling, and failover processes.
- Collaborate with AI Risk Labs and the insurance team to translate risk methodology into model performance.
Qualifications
- Strong foundation in Machine Learning and Deep Learning fundamentals.
- Hands-on experience in fine-tuning Large Language Models (LLMs) or foundation models using techniques such as LoRA/QLoRA.
- Proficiency in PyTorch and the Transformers ecosystem.
- Skills in quantization, hyperparameter tuning, and inference optimization (e.g., Speculative Decoding).
- Experience with Distributed Training methodologies (such as DeepSpeed, FSDP, Accelerate).
- Ability to set up and operate training/severing infrastructure, including GPU provisioning and containerization with frameworks like SGLang or vLLM.
- Bachelor's degree or higher in Computer Science, Engineering, or a related field.
Preferred Skills
- Experience in training models from scratch or working with models containing 2B+ parameters.
- Familiarity with cloud inference at scale, particularly with AWS Bedrock and SageMaker, as well as managing quota and throttling.
- Good understanding of AI security or adversarial machine learning concepts (e.g., using Promptfoo, Garak, Giskard, or similar tools).
- Exposure to regulated or compliance-heavy environments such as security, fintech, or insurance sectors.
Experience
- 3 to 5 years of relevant experience in AI/ML engineering or related fields, with a focus on foundation models and infrastructure.
Environment
This position is onsite in Bangalore. Specific details about the work environment or additional physical conditions were not specified.
Salary
Salary information was not provided.
Growth Opportunities
Details on career advancement opportunities within the company are not specified.
Benefits
A comprehensive list of offered benefits is not outlined in the description.
