Artificial Intelligence Engineer

Artificial Intelligence Engineer - Foundation Models & Scientific Machine Learning
Location: Remote
Engagement: Full-Time
Experience: 4–8+ Years
Compensation: Open to discussion
About the opportunity
We are an early-stage, US-based deep technology company developing proprietary AI foundation models and advanced predictive systems at the intersection of artificial intelligence, scientific computing, and complex data intelligence.
We are seeking a highly skilled, hands-on AI Engineer to join our India-based engineering team and contribute to the research, development, training, and deployment of next-generation AI architectures.
This role requires strong fundamentals in deep learning, mathematical modeling, and scalable ML systems, with the ability to translate cutting-edge research into production-ready solutions.
Key Responsibilities
1. Foundation Models & Advanced Deep Learning
Design, implement, train, and optimize neural architectures, including Graph Neural Networks (GNNs), Graph Transformers, CNNs, and attention-based Transformers.
Develop multimodal representation learning frameworks integrating structured, unstructured, relational, and high-dimensional datasets.
Implement self-supervised, contrastive, transfer, and multi-task learning approaches.
Explore geometric deep learning, equivariant neural networks, and domain-specific foundation model architectures.
2. Generative AI & Scientific Modeling
Research and implement generative architectures, including diffusion models, score-based generative models, VAEs, and generative Transformers.
Develop predictive and generative models for complex structured data, graphs, and continuous representations.
Work with sparse, noisy, and heterogeneous datasets, addressing generalization, uncertainty, and data limitations.
Explore probabilistic modeling, causal inference, and uncertainty-aware machine learning.
3. Knowledge Graphs, RAG & Reasoning Systems
Design knowledge graph architectures and graph-based representation learning pipelines.
Develop Retrieval-Augmented Generation (RAG), GraphRAG, hybrid retrieval, semantic search, and reranking systems.
Integrate LLMs, embeddings, vector databases, and structured knowledge for evidence-grounded reasoning.
Implement retrieval evaluation, source attribution, factual consistency checks, and hallucination mitigation.
4. Model Training, Evaluation & Optimization
Build scalable training pipelines using PyTorch and GPU-accelerated computing.
Implement custom loss functions, optimization techniques, and distributed training strategies.
Establish rigorous model evaluation frameworks, including ablation studies, cross-validation, calibration, and out-of-distribution testing.
Optimize model performance, memory utilization, training efficiency, and inference latency.
5. AI Infrastructure & Research Engineering
Develop reproducible ML pipelines, experiment tracking, model versioning, and production inference systems.
Build scalable model-serving APIs and integrate trained models into larger AI platforms.
Implement research papers, benchmark alternative architectures, and document experimental findings.
Collaborate with interdisciplinary researchers and engineers to develop proprietary AI capabilities.
Required Qualifications
4–8+ years of hands-on experience in AI/ML engineering, deep learning, or applied AI research.
Advanced proficiency in Python and PyTorch, with strong software engineering fundamentals.
Solid experience implementing and training deep learning architectures, particularly GNNs, Transformers, or CNNs.
Strong understanding of representation learning, neural network optimization, and model generalization.
Experience with large-scale datasets, GPU-based model training, and ML experimentation.
Strong mathematical foundations in linear algebra, probability, statistics, and numerical optimization.
Experience with model evaluation, benchmarking, debugging, and performance optimization.
Familiarity with cloud infrastructure, Docker, Git, and production ML workflows.
Ability to independently understand and implement research papers.
Preferred Qualifications
Experience with geometric deep learning, equivariant GNNs, Graph Transformers, or 3D representation learning.
Experience with diffusion models, score-based generative modeling, or other advanced generative architectures.
Hands-on experience with RAG, GraphRAG, knowledge graphs, vector databases, or neural-symbolic AI.
Knowledge of multimodal foundation models, cross-modal attention, and domain adaptation.
Experience with causal inference, Bayesian methods, uncertainty quantification, or probabilistic modeling.
Familiarity with PyTorch Geometric, DGL, Hugging Face, DeepSpeed, FSDP, or equivalent frameworks.
Exposure to scientific machine learning, computational biology, molecular modeling, or other research-intensive domains.
Research publications, open-source contributions, or demonstrated experience developing original model architectures.
We are looking for an engineer who can:
Design and develop original AI models beyond standard API integrations.
Understand the mathematical foundations and architectural trade-offs behind deep learning systems.
Work independently on complex research and engineering challenges.
Translate scientific research into robust, scalable AI implementations.
Take ownership of the complete model lifecycle, from experimentation and validation to deployment.
This is a core AI engineering role focused on building proprietary predictive intelligence and foundation model infrastructure, not primarily prompt engineering or LLM application development.
Engagement and Compensation
Full-time, remote opportunity with our engineering team collaborating with international technical and research leadership.
Compensation: Open to discussion based on experience, technical expertise, and scope.
To Apply: Please share your resume, GitHub profile, and examples of original AI/ML models, research implementations, or complex systems you have developed.

Work arrangement
Yes