Research Engineer, Applied AI

Akoncagua AI; Lakeland (FL); On-site; USD 120,000 - 180,000; Full time

Job summary

Akoncagua AI is seeking exceptional Research Engineers who combine deep machine learning expertise with strong software engineering fundamentals. This role covers the full model development lifecycle—from architecture experimentation and data curation to training, evaluation, optimization, and deployment.

You will work closely with researchers and engineers to develop state-of-the-art language and multimodal AI systems while contributing to the technical direction of next-generation foundation

Qualifications

• MS, PhD, or equivalent industry experience in a quantitative field.
• Experience training and fine-tuning large language models (distillation, continual pre-training, supervised fine-tuning, policy optimization).
• Deep understanding of Deep Learning and transformer architectures; designing or modifying them.
• Proficiency in Python and PyTorch.
• Ability to read, evaluate, and adapt current research literature.
• Strong software engineering skills in algorithms, data structures, debugging, profiling, and systems design.

Responsibilities

• Develop and optimize language and multimodal AI systems across the model lifecycle.
• Experiment with architectures, data curation, training, evaluation, and deployment.
• Collaborate with researchers and engineers to translate cutting-edge research into production-ready systems.

Skills

Deep Learning

Transformer architectures

Software engineering

Research ability

Distributed systems

Education

MS/PhD or equivalent in a quantitative field

Tools

Python

PyTorch

Job description

Akoncagua AI is building domain-specialized foundation models for technically demanding industries. Our work spans large language models, multimodal systems, synthetic data generation, alignment, evaluation, and scalable training infrastructure. We are seeking exceptional Research Engineers who combine deep machine learning expertise with strong software engineering fundamentals. This role is ideal for individuals who enjoy working across the entire model development lifecycle—from architecture experimentation and data curation to training, evaluation, optimization, and deployment. You will work closely with researchers and engineers to develop state-of-the‑art language and multimodal AI systems while contributing to the technical direction of next‑generation foundation models.

Role Overview

This role sits at the intersection of machine learning research, systems engineering, and product development. The ideal candidate possesses strong theoretical foundations in deep learning alongside practical experience building, training, adapting, evaluating, and optimizing modern AI systems at scale.

Successful candidates are comfortable working across model architecture, training methodologies, alignment techniques, multimodal learning, evaluation frameworks, and distributed infrastructure while translating cutting‑edge research into production‑ready systems.

Required Qualifications

• MS, PhD, or equivalent industry experience in Computer Science, Artificial Intelligence, Machine Learning, Applied Mathematics, or a related quantitative field.
• Experience in methods of training and fine‑tuning large language models, such as distillation, continual pre‑training, supervised fine‑tuning, and policy optimization
• Deep understanding of Deep Learning, transformer architectures and experience in designing or modifying them.
• Proficiency in Python and PyTorch.
• Demonstrated ability to read, evaluate, and adapt techniques from current research literature.
• Strong software engineering skills with demonstrated proficiency in algorithms, data structures, debugging, profiling, and systems design.

Preferred Qualifications

• Strong publications (NeurIPS, CVPR, ICML, ICLR, ICCV, ACL), patents, or open source contributions in related areas.
• Familiarity with model architecture research, attention variants, efficient transformers, mixture‑of‑experts systems, retrieval‑augmented architectures, or agentic systems.
• Experience with inference and serving optimization, including: Speculative Decoding, KV Cache Optimization
• Demonstrated experience designing datasets and building synthetic datasets for specialized or low‑resource domains.
• Experience with the Hugging Face ecosystem, including Transformers, Datasets, Accelerate, PEFT, and TRL.
• Familiarity with large‑scale training infrastructure using DeepSpeed, FSDP, DDP, ZeRO optimization and Distributed GPU clusters.