Job Purpose
Design and deliver robust AI solutions for defined use cases, addressing complex (non-trivial) modeling, evaluation, and production challenges within project scope.
Apply and extend existing patterns, tools, and frameworks to improve AI system quality, performance, and efficiency within the team.
Ensure reliable end-to-end AI lifecycle execution (data, model, deployment) in production environments, following established standards and practices.
Act as a strong technical contributor, resolving complex issues within scope, supporting team delivery, and operating effectively under real-world constraints with moderate guidance.
Key Accountabilities (1)
AI Craft - Specialization
Evaluate novel algorithms and training techniques against production constraints; diagnose and unblock the hardest cross-cutting Modeling problems (severe distribution shift, low-resource settings, multi-task architecture); build reusable Modeling patterns adopted across teams.
Implement complex multi-agent orchestration, long-running stateful workflows, and quality optimisation at scale; evaluate new models and frameworks hands-on; build reusable agent patterns and prompt libraries adopted across teams.
Evaluates and adopts advanced optimisation techniques; designs multi-tenant serving and capacity strategies; builds performance-regression frameworks and cost dashboards adopted across teams.
Recognised expert in the sub-domain inside the org; selects between competing approaches based on production constraints; reviews and unblocks others' sub-domain work; tackles the harder problems (low-resource scenarios, distribution shift, real-time constraints); the person teams escalate to.
AI Craft - Evaluation
Build evaluation tooling and automated experiment pipelines adopted across teams
Design and run online A/B tests with statistical rigour & calibrate human and LLM-judge evaluation
Resolve the hardest evaluation problems (subjective tasks, rare-event detection, long-tail safety).
Key Accountabilities (2)
Engineering for Production
Design testing strategies for AI systems (data validation, model regression, prompt-output, agent-trajectory tests); evaluate and introduce new tooling; build internal libraries, SDKs, and code templates that raise the engineering floor across teams.
Architect data platforms that serve both ML training and LLM/RAG workloads; design for cost, latency, and freshness trade-offs; build reusable data-quality, lineage, and observability tooling adopted across teams.
Architect complex AI systems combining ML and agentic components; design for graceful degradation, cost-efficiency, evolvability, and multi-tenancy; build reference implementations and integration patterns adopted by other teams; lead migration and convergence of inherited or siloed systems where warranted; balance near-term delivery against long-term architectural coherence.
Success Profile - Qualification and Experiences
Qualifications
Bachelor's or master's degree in computer science, Artificial Intelligence, Data Science,
Software Engineering, Information Technology, or a related quantitative field
Strong foundation in mathematics, statistics, and machine learning principles
Relevant certifications in AI/ML, Cloud (Azure, AWS, GCP), or Data Engineering are a plus
English proficiency in line with Techcombank's policy
Work Experience
2-3+ years of experience in AI/ML, data engineering, and software development, with proven delivery of end-to-end AI solutions in production environments
Strong expertise in AI specialization, including machine learning, deep learning, NLP, and LLM applications, RAG with hands-on experience in experimentation, model evaluation, and performance tuning
Demonstrated experience in designing and building scalable AI systems and architectures, including microservices-based solutions and distributed systems
Solid experience in data engineering for AI, including building ETL/ELT pipelines, data lakes/warehouses, and handling structured & unstructured data using tools like Spark, Kafka, Airflow, and SQL/NoSQL databases
Proficiency in programming languages such as Python, Go, Java, or Scala, and experience with AI/ML frameworks (e.g., TensorFlow, PyTorch, / JAX, transformers, fine-tuning (LoRA, DPO), distillation, evaluation.)
Experience in AI experimentation & evaluation, including A/B testing, model benchmarking, validation techniques, and monitoring model performance in production
Working knowledge of AI Safety, Ethics & Responsible AI, Red-teaming, model risk management, alignment with SBV, MAS, Basel III, ensuring fairness, transparency, explainability, and compliance with regulatory requirements
Experience in MLOps/LLMOps practices (Level 2), including model deployment, CI/CD pipelines, versioning, and monitoring (e.g., Kubernetes, GPU orchestration, vector DBs, CI/CD for ML, drift and bias monitoring.)
Proven ability in business needs analysis, translating business problems into AI-driven solutions and measurable outcomes
Experience managing stakeholders, collaborating with cross-functional teams (business, risk, IT), and influencing decision-making
Track record of delivering AI projects end-to-end using Agile methodologies, ensuring timelines, quality, and business impact
Exposure to emerging technologies and research trends in AI, with the ability to evaluate and adopt new approaches when relevant