Software Engineer II, Analytics Compute Platform Team

Amplitude; San Francisco (CA); Hybrid; USD 140,000 - 190,000; Full time

Job summary

In a fast-changing system, you’ll own small-to-medium projects end to end, learn from strong engineers, and contribute to a culture of learning and growth. This SF-based role offers a hybrid work model in a growing company that serves thousands of customers.

Qualifications

• 2–4 years of experience building and running production services.
• Write clear, tested code and reason about failure, retries, and correctness in data-intensive systems.
• Instrument code, verify outcomes, and be proactive in quality.

Responsibilities

• Design and ship new query capabilities for behavioral insights.
• Own reliability: define tests, rollout plans, observability, and SLOs.
• Keep real customer workloads fast by profiling and optimizing query paths.
• Participate in migrations of APIs, schemas, and services with safe rollouts.
• Build deep system knowledge and share learnings through docs and demos.
• Stay close to customer problems and measure impact of your work.

Skills

Backend engineering

Problem solving

Ownership of quality

Clear communication

Customer curiosity

AI-assisted engineering

Learning and growth

Job description

Amplitude is the leading AI analytics platform, helping over 5,200 customers—including Atlassian, Burger King, NBCUniversal, and Square—build better products and digital experiences. With powerful AI Agents embedded across our platform, teams can analyze, test, and optimize user experiences faster than ever. Ranked #1 across multiple categories in G2’s Winter 2026 Report, Amplitude is the best-in‑class solution for product, data, and marketing teams. Learn more at amplitude.com.

As an organization, we deliver for our customers by living our values. We operate from a place of humility, take ownership of problems and successes, approach challenges with a growth mindset, and put our customers at the center of everything we do.

San Francisco- Hybrid

ABOUT THE ROLE & TEAM:

Every chart, insight, and experiment result a customer sees in Amplitude passes through the compute layer our team owns. We sit between Amplitude’s product experiences, APIs, and MCPs and our analytics database, and we own the compute API that turns a user’s question into a fast, correct answer.

Our mandate: enable a highly performant, reliable, and flexible way to compute insights. Day to day, you’ll work with engineers on Data Management, marketing analytics, Session Replay, Experiment, CDP, and other product teams that build on what we ship, as well as with the Query team below us, whose engine we rely on.

This role is for an engineer who is hungry to grow quickly. You’ll own small-medium-complexity projects end to end in a system that is changing fast, and we’ll stretch you on purpose with real problems, honest feedback, and strong engineers to learn from.

WHAT YOU’LL ACHIEVE:

• Make new behavioral insights easier to build. Design and ship new query capabilities, extension points, and shared libraries/services that let product teams deliver new analyses in less time.

• Own the reliability of what you build. Define tests, rollout plans, and observability, know your service’s SLOs, and fix problems at the root.

• Keep real customer workloads fast. Profile query paths, find bottlenecks with traces and experiments, and verify that your changes helped.

• Help us change the system safely. Take part in migrations of APIs, schemas, and services, and learn how good abstractions make change cheaper.

• Become a go-to person. Build real expertise in part of the system, answer teammates’ questions, and share what you learn through docs and demos.

• Stay close to the customer. Understand the customer problem behind every project, measure whether your work made their experience better, and bring what you learn back to the team.

EXAMPLE PROJECTS YOU MIGHT WORK ON:

• Extend the compute API so a new kind of behavioral question can be expressed.

• Propose and Move a set of query paths onto a new internal representation, with parity testing and a safe rollout so that future improvements can be done safely.

• Add observability and SLO tracking to a part of the compute layer that doesn’t have it yet.

• Profile and speed up a slow query pattern reported by a product team.

SKILLS YOU’LL NEED TO BRING:

• Backend engineering fundamentals: You have 2–4 years of experience building and running production services. You write clear, tested code and can reason about failure, retries, and correctness in data-intensive systems.

• Thoughtful problem‑solving: You start by understanding the context, decompose tricky problems, and work toward a clean solution. You ask clarifying questions before diving in.

• Ownership of quality: You instrument your code, verify outcomes before calling work done, and respond quickly when something you shipped breaks.

• Clear communication: You flag risks early, explain decisions in writing and in conversation, and give and receive feedback openly.

• Customer curiosity: You ask what problem a customer is trying to solve before you start building. You use Amplitude yourself, can explain it to others, and care whether the feature you shipped made someone’s work better.

• Fluency with AI‑assisted engineering: You use AI to speed up learning, exploration, and testing while taking responsibility for understanding and validating the result.

• A habit of learning and improving: You seek out feedback, act on it, and can point to specific things you got better at recently. You’re curious about how the system works around the layers you’re working on.

NICE TO HAVES:

• Experience with query engines, columnar or analytical databases, or large-scale data processing.

• Experience with observability, SLOs, and oncall for a production service.

• Side projects, writing, or talks that show you learn in public.

HOW WE WORK:

• Ship fast. We create clarity in ambiguity through action. We favor prototypes, MVPs, and incremental milestones over long planning cycles, and we elevate when a decision is stuck.

• Value impact over effort. We tie every initiative to team, organization, and company goals, and we reshape or stop work when the expected impact no longer justifies it.

• Company > team > me. We make decisions for the whole organization, share credit generously, and invest in the growth of the engineers around us.