Twilio
Machine Learning Engineer 2
$129K–$170K
Remote - US · Remote · Full-time · Posted Oct 28, 2026
About the job
Build and operate pipelines and evaluation systems that extract, retrieve, and measure accurate customer memory from conversation data for Twilio’s Conversational AI platform.
What you'll do
- Build extraction pipelines and tune LLM prompts and schemas to create structured customer memory.
- Own memory retrieval using embeddings, vector search, and ranking, and contribute to quality benchmarks and regression gates.
- Diagnose production issues using logs, metrics, and traces; partner with Product and ML teams on system changes and evaluation criteria.
What they ask for
- Experience building backend services and ability to learn new languages, frameworks, and tools.
- Experience building or working with evaluation or quality-measurement systems, or desire to learn.
- Grounding in observability practices, including structured logging and metrics, and using them to debug distributed systems in production.
- Fluency with AI-assisted development tools, applied with deliberate direction and critical thinking.
Nice to have
- Experience with Go and/or Python.
- Experience with event-streaming systems such as Kafka, LLM-backed products, or practical ML.
Pay
$128,560 - $160,700 (base salary); $136,000 - $170,000 (base salary)
Tools
LLM, embeddings, vector search, Go, Python, Kafka
AI focus
Build LLM-based extraction and retrieval systems for customer memory, tune prompts and schemas, and evaluate memory quality; use AI-assisted development tools.
Model-written summary · the original posting is authoritative