Lead SLM/VLM Engineer — Edge AI & Efficient Inference

Jabil; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Jabil is seeking a highly capable engineer/researcher to lead the R&D of Small Language Models and Vision-Language Models for edge/low-latency production scenarios.

You will own the continuous pretraining, supervised instruction tuning (SFT), and compression/distillation pipelines, and work closely with platform teams to deliver reliable, measurable improvements in inference efficiency, tool-use success rate, and overall model quality.

Qualifications

• Strong software engineering skills in Python and C++, including experience building ML training/evaluation pipelines in PyTorch.
• Hands-on experience in model efficiency and inference optimization (e.g., distillation, quantization, pruning, serving optimization).
• Experience with high-performance computing and acceleration: CUDA and/or SIMD, profiling and performance tuning.
• Ability to read and reproduce key ideas from recent papers and implement algorithms with strong experimental discipline.
• Ability to communicate effectively in both Chinese (Mandarin) and English as the successful candidate will have to liaise with our counterparts in China.

Responsibilities

• SLM/VLM Training: Continuous Pretraining & Instruction Tuning (SFT) and build reproducible training workflows in PyTorch.
• Compression, Distillation & Edge/Low-Latency Inference Optimization: implement efficient compression strategies and optimize throughput and cost-per-token.
• Tool Calling System: Architect and implement a production-grade tool calling framework with cataloging, routing, validation, and observability.
• RL & Reward Modeling for Alignment and Tool-Use Reliability: apply PPO/DPO/GRPO-like methods and reward modeling for alignment and reliability.
• Data Pipeline Automation: design automated pipelines for data collection, cleaning, labeling, and dataset version management.
• Rigorous Evaluation, Testing & Iteration: offline benchmarks, regression tests, and reliability metrics with A/B tests and ablations.

Skills

Python

C++

PyTorch

Model optimization

Distillation

Tools

CUDA

Job description

Jabil is seeking a highly capable engineer/researcher to lead the R&D of Small Language Models and Vision-Language Models for edge/low-latency production scenarios.

You will own the continuous pretraining, supervised instruction tuning (SFT), and compression/distillation pipelines, and work closely with platform teams to deliver reliable, measurable improvements in inference efficiency, tool-use success rate, and overall model quality.