Junior AI Engineer

About ATNS

At ATNS, we, together, design and deliver end-to-end technology solutions that are driven by a product mindset and a deep understanding of business needs. Every solution we build is shaped by our client goals, our customer-first approach lets them focus on what matters while we, together, power their digital transformation through product-driven solutions.

Beyond solutions, ATNS is passionate about our people . We are dedicated to building a Vietnamese workforce that meets international standards, empowering local talent to shine on the global stage. By investing in our people and fostering learning and development, we aim to be the trusted partner that helps organizations shape a strong, sustainable future growth.

Job Description

Join a high-impact AI Logistics initiative, supporting the digital transformation of logistics operations through advanced AI agent integration and data engineering. This project is mission-critical, with a focus on enabling seamless data-to-agent workflows and empowering the local technical team for long-term success.

Role Overview

The Junior AI Engineer will participate in developing and deploying Artificial Intelligence (AI) solutions within projects.

• This role focuses on supporting the building, training, and optimization of AI models, while collaborating with data engineers and data scientists to bring models into real world applications.
• This is an entry-level position suitable for recent graduates or candidates with basic experience who are eager to grow their career in AI.
  
  

Your Responsibilities

• Assist in designing, building, and testing AI agent workflows (tool use, multi-step reasoning, retrieval-augmented generation).
• Build and maintain secure integrations between agents and external tools, APIs, and data sources.
• Implement data preprocessing and context pipelines that feed agents clean, well structured input.
• Instrument agent workflows with logging, tracing, and evaluation; debug agent behavior from traces and reproduce failures systematically.
• Write and optimize clean, maintainable code following standard agent design patterns and engineering best practices.
• Test, evaluate, and improve agent reliability, cost, and latency against real-world data.
• Collaborate with data engineers and senior engineers to deploy agents into production environments.
• Research and stay updated on new agent engineering methodologies, models, and orchestration patterns.
  
  

Success Metrics (first 6–12 months)

• Study: Be open-minded and continuously improve your skillset through internal and external learning.
• Data-to-Agent Preparation: All required data pipelines and agent workflows are production-ready.
• Testing Completion: Agent-level testing, tracing, and debugging completed before go-live.
• Production Readiness: Deployed agents demonstrate robust performance, reliability, and scalability, with observability in place.
• Local Team Enablement: Local technical team is trained and capable of independent operation..
  
  

Qualifications

• Must‑have
• Proficiency in Python or JavaScript — strong scripting and automation skills for building agent workflows, integrations, and backend services.
• Hands-on experience building LLM agent workflows — practical experience applying standard agent engineering practices: tool/function calling, structured output handling, prompt design, and agent state management (via any orchestration framework or custom harness code). Personal projects, internships, and open-source contributions all count.
• Understanding of LLM mechanics — context windows, token limits, prompt engineering, and structured output parsing (JSON mode, function calling, tool schemas).
• Agent orchestration fundamentals — familiarity with standard methodologies such as ReAct loops, tool-use patterns, memory management (short-term / long-term), and agent state machines.
• API integration & harnessing — experience wiring agents to external tools, APIs, and data sources securely and reliably, with basic security hygiene (secrets handling, least privilege access, prompt-injection awareness).
• Data handling basics — ability to preprocess, structure, and pipeline data for agent context injection and retrieval-augmented generation (RAG).
• Testing & debugging discipline — ability to write tests for non-deterministic agent behavior, read logs and traces, and reproduce failures methodically.
• Version control & collaboration — Git, code review, and working in agile teams.
• Strong English communication — ability to articulate technical designs, debug logs, and agent behavior clearly in an international environment.
  

Nice-to ‑ have

• Production agent deployment — experience deploying agents with observability, tracing, and error-handling guardrails following standard LLMOps practices.
• Agent evaluation — building evaluation sets or golden datasets, LLM-as-judge scoring, and regression testing of prompts and agent strategies.
• Cost & latency optimization — awareness of token budgeting, caching strategies, and model selection / routing trade-offs.
• Retrieval tooling — exposure to vector stores, chunking, and embedding strategies for retrieval-augmented generation.
• Multi-agent system design — understanding of agent-to-agent communication, delegation patterns, and conflict resolution in distributed agent networks.
• Programming paradigm breadth — awareness of different programming paradigms (object-oriented, functional, procedural, event-driven) and when to apply each; exposure to functional programming concepts such as immutability, pure functions, and composition is a plus.
• Cloud & infrastructure exposure — familiarity with cloud services for hosting agent runtimes, containerization, and basic CI/CD.
• MLOps / LLMOps practices — model versioning, prompt versioning, A/B testing of agent strategies, and performance monitoring.
• Logistics or supply chain domain knowledge — prior exposure to route optimization, inventory systems, or operational planning workflows.
• Multicultural team experience — comfortable working across time zones with distributed engineering and product teams.
  
  

Education

• Bachelor’s degree in Computer Science, IT, Mathematics, Physics, or a related field — or equivalent demonstrated experience via projects / open-source contributions.
  
  

Benefits

BENEFITS DESCRIPTION Meal allowance Monthly 880,000 VND Extra Insurance_Employee Included Annual Health Check Included 13th month salary Included Performance Bonus 0-6 salary Working type Hybrid (3 days office, 2 wfh), no check in – check out Working days Sunday – Thursday Annual Leave Base of 12 days plus add 01 day off for every 3 years working at the company. Team building Included Annual vacation, travel Included Training/Course Included Pantry/Happy Hour/ Birthdays Included Certificate Included Study Partially or fully paid master/specialization degree Year End Party Included Staff Awards Included Holidays/Festivals Included Welcome Onboard Gift Kit Included

SALARY STRUCTURE

Cost Note Monthly Salary Paid first day every month Quarterly Bonus Paid in the first month of the next Quarter 13th Salary Paid before Tet Yearly Bonus Paid before Tet

CAREER PATH

Learn How This Role Can Grow Within Our Company

Career_Development_Plan_AI_Junior