Machine Learning Engineer 2

Twilio; San Francisco (CA); Remote; USD 136,000 - 170,000; Full time

Benefits offered by this job

Health insurance

401(k)

Paid sick time

Paid personal time off

Paid parental leave

Job summary

Twilio Inc. is seeking a Machine Learning Engineer for Customer Memory to build and scale pipelines that convert raw conversation data into structured memory. You will ensure accurate memory as the product evolves, and you’ll work on the memory layer behind Twilio’s Conversational AI platform.

You’ll focus on high-quality extraction, fast retrieval, and robust evaluation. You’ll collaborate with Product and ML teams to translate customer problems into system changes, and you’ll contribute to

Qualifications

• Backend services experience with production systems.
• Curious about service management, observability, and reliability.
• Experience with evaluation/quality metrics in ML/AI.
• Familiarity with AI-assisted dev tools and code generation.
• Willingness to learn new languages and frameworks.

Responsibilities

• Build pipelines to convert raw data into structured memory.
• Own retrieval path using embeddings, vector search, and ranking.
• Contribute evaluation systems with quality metrics and benchmarks.
• Diagnose production issues across ingestion, extraction, and indexing.
• Collaborate with Product and ML teams to translate problems into changes.

Skills

Backend development

Observability

Quality metrics

AI tooling

Go/Python

Kafka

LLM integration

ML deployment

Tools

Kafka

Job description

See yourself at Twilio

Join the team as Twilio's next Machine Learning Engineer, Customer Memory.

About the job

This position is needed to build and scale the pipelines that turn raw conversation data into structured, searchable customer memory, and the evaluation systems that keep that memory accurate as the product evolves.

You'll join the Conversation Memory team, which builds the memory layer behind Twilio's Conversational AI platform: the systems that extract what matters from millions of conversations, resolve it to the right customer, and retrieve it in milliseconds when an agent needs it. Memory that is wrong, stale, or crossed between customers is worse than no memory at all, so quality measurement is a first-class part of the work.

Responsibilities

In this role, you'll:

• Build the extraction pipeline that turns raw data into structured memory, tuning the LLM prompts and schemas that decide what an AI agent remembers about a customer and what it forgets.
• Own the retrieval path that answers "what do we know about this person?" in milliseconds, working with embeddings, vector search, and ranking
• Contribute to the eval systems that tell us whether memory is actually getting better, turning a fuzzy question of quality into benchmarks and regression gates that run before every change ships.
• Diagnose production issues across the ingestion, extraction, and indexing path using logs, metrics, and traces.
• Partner with Product and ML teams to translate customer problems into system changes and evaluation criteria.

Qualifications

Twilio values diverse experiences from all kinds of industries, and we encourage everyone who meets the required qualifications to apply. If your career is just starting or hasn't followed a traditional path, don't let that stop you from considering Twilio. We are always looking for people who will bring something new to the table!

Required:

• Experience building backend services, and a demonstrated ability to pick up new languages, frameworks, and tools as the problem demands.
• A curious approach to service management: ability to investigate how systems actually behave in production, and desire to continuously learn how to operate them with rigor and accountability.
• Experience building or working with evaluation/quality-measurement systems, such as regression detection or quality metrics, or the desire to learn.
• Solid grounding in observability practices: structured logging, metrics, and using them to debug distributed systems in production.
• Fluency with AI-assisted dev tools (e.g. agents for code generation, review, debugging), applied with deliberate direction and critical thinking.

Desired:

• Experience with Go and/or Python.
• Experience with event-streaming systems (e.g. Kafka).
• Experience building LLM-backed products.
• Practical experience applying ML, e.g. building, evaluating, or productionizing models.

Location

• This role will be remote, but is not eligible to be hired in San Francisco, CA, Oakland, CA, San Jose, CA, or the surrounding areas.

Travel

We prioritize connection and opportunities to build relationships with our customers and each other. For this role, you may be required to travel occasionally to participate in project or team in-person meetings.

What We Offer

Working at Twilio offers many benefits, including competitive pay, generous time off, ample parental and wellness leave, healthcare, a retirement savings program, and much more. Offerings vary by location.

Compensation

*Please note the salary range information provided applies only to candidates residing in California, Colorado, Hawaii, Illinois, Maryland, Massachusetts, Minnesota, New Jersey, New York, Vermont, Washington D.C., and Washington State due to local requirements. Compensation for candidates in other locations will be discussed during the hiring process. Please note that hiring for this role is not restricted to the locations listed above.

The estimated pay ranges for this role are as follows:

• Based in Colorado, Hawaii, Illinois, Maryland, Massachusetts, Minnesota, Vermont or Washington D.C. : $128,560 - $160,700
• Based in New York, New Jersey, Washington State, or California (outside of the San Francisco Bay area): $136,000 - $170,000.
• This role may be eligible to participate in Twilio's equity plan and corporate bonus plan. All roles are generally eligible for the following benefits: health care insurance, 401(k) retirement account, paid sick time, paid personal time off, paid parental leave.

The successful candidate's starting salary will be determined based on permissible, non-discriminatory factors such as skills, experience, and geographic location.

Application deadline information

Applications for this role are intended to be accepted until November 13, but may change based on business needs.