AI Research Engineer

ABOUT CLIENT

Our client is a leading research company specializing in technology innovation

JOB DESCRIPTION

Focus on improving model usability for users
Conduct training runs and AI experiments
Analyze results and make necessary changes
Collaborate with the product engineering team to implement improvements
Implement and improve upon recent RL techniques like GRPO, DPO, RePO, etc.
Create and manage adaptable, expandable training codebases
Establish and maintain efficient data pipelines, including both synthetic and real data
Ensure training jobs are capable of scaling across multiple GPUs and nodes, such as FSDP, DDP, NCCL
Maintain code health over the long term by writing clean, testable, and reproducible code
Contribute to the enhancement of open source dependencies
(Optional) Publish papers and present research findings

JOB REQUIREMENT

Comprehensive skills in Python and coding using frameworks like PyTorch, or similar
Demonstrated proficiency in training deep learning and reinforcement learning models in practical scenarios
Experience in handling and analyzing large datasets and intricate workflows
Adept understanding of training dynamics, identifying issues and troubleshooting them
Proficiency in job launchers, logging tools (such as Weights & Biases, TensorBoard), and checkpointing systems
An approach of applying engineering precision to research through writing clear code, meticulous design, and reproducible results
Knowledge of TorchScript, ONNX, or custom inference runtimes
Contribution to open-source projects related to PyTorch or machine learning tools
Background in working with transformer models, diffusion models, VLMs, or extensive vision/NLP tasks
Familiarity with batch schedulers (SLURM), cluster environments, and GPU resource management
Ability to collaborate closely with systems engineers or MLOps teams for seamless integration

WHAT'S ON OFFER

Join an exceptional research team to work on significant and impactful projects
Take charge of and influence the primary training code infrastructure utilized by the team
Engage with actual models, real data, and substantial scale challenges, not small-scale problems
Contribute to bridging the gap between research speed and engineering excellence
Enjoy a flexible work setting with a culture that treasures depth, transparency, and inquisitiveness

CONTACT

PEGASI – IT Recruitment Consultancy | Email: recruit@pegasi.com.vn | Tel: +84 28 3622 8666
We are PEGASI – IT Recruitment Consultancy in Vietnam. If you are looking for new opportunity for your career path, kindly visit our website www.pegasi.com.vn for your reference. Thank you!

Job Summary

Company Type:

Product

Technical Skills:

Machine Learning, Python, AI

Location:

Ho Chi Minh - Viet Nam

Working Policy:

Onsite

Salary:

Negotiation

Job ID:

J01951

Status:

Close

Related Job:

PreSales Solutions Engineer

Ho Chi Minh - Viet Nam

Product

  • Presale
  • System
  • Google Cloud

PreSales Support: Collaborating with the Sales team to understand client needs and develop tailored solutions using Google Maps and Google Cloud services. This involves conducting technical presentations, product demonstrations, and creating proof of concepts (POCs) for prospective clients, as well as contributing to proposals and RFP responses with detailed technical information. Post-Sales Support: Leading the technical implementation of Google Maps and Google Cloud services, ensuring smooth deployment and integration. Providing ongoing technical support and troubleshooting for clients after implementation, working closely with cross-functional teams to ensure client satisfaction and build long-term relationships. Technical Expertise: Staying up-to-date with the latest Google Maps and Google Cloud technologies, serving as a subject matter expert (SME) for both internal teams and clients. Integrating new features and services into client solutions and providing guidance on best practices. Collaboration: Working closely with Sales, Product, Infrastructure, Data, and Engineering teams to align solutions with client needs and company goals. Mentoring junior team members and contributing to training initiatives.

Negotiation

View details

Technical Lead

Ho Chi Minh - Viet Nam

Product

  • NodeJS
  • Python

Leading the backend development team, providing technical direction, mentorship, and best practices. Designing and implementing scalable, secure, and high-performance microservices-based architectures. Architecting and implementing agentic AI workflows and RAG (Retrieval-Augmented Generation) systems for personalized user interactions and automated coaching features. Overseeing data pipelines and infrastructure required for real-time AI model inference within a microservices-based environment. Collaborating with stakeholders to align on requirements and delivery timelines. Optimizing application performance, monitoring system reliability, and proactively troubleshooting issues. Advocating for CI/CD pipelines, automated testing, and robust version control strategies. Documenting key architectural decisions, APIs, and processes for internal use.

Negotiation

View details

Chief Technology Officer

Ha Noi - Viet Nam

Product

  • Cloud
  • Backend

Planning & designing overall system architecture: Creating a Technology Roadmap for a Game Server system with high concurrency and low latency for global players. Cost optimization: Deciding on the strategy for using Cloud infrastructure (AWS, GCP, Azure) or Hybrid Cloud to balance performance and operational expenses. High-level consultation: Participating in the Executive Board to address the relationship between speed-to-market of features and system stability. Tech-stack selection: Evaluating and finalizing programming languages (Go, C++, Java, Node.js) and processing models (Microservices vs Monolith) suitable for the complex logic of the game. Scalability solution: Directing the development of Auto-scaling, Load Balancing mechanisms, and managing Player State on large clusters. Data management: Designing Database structure (SQL/NoSQL) and Cache system (Redis, Memcached) to handle billions of queries daily without congestion. Ensuring Uptime: Building real-time monitoring and alerting systems to maintain 99.99% Availability. Network security: Implementing solutions to combat DDoS attacks, game fraud (Anti-cheat), and comprehensive user data security. Infrastructure & CI/CD: Standardizing automatic deployment processes to ensure game updates (Hotfix/Update) do not disrupt players. Deployment strategy & Optimization: Developing plans to optimize Cloud Services costs (AWS/GCP/Azure), evaluating the use of Spot Instances, Reserved Instances, or Private Cloud solutions to save operational budget. Meanwhile, establishing 24/7 monitoring and incident response systems.

Negotiation

View details