Software Engineer

Harmony (tryharmony.ai); United States; On-site; USD 112,500 - 137,500; Full time

Benefits offered by this job

Competitive salary

Equity opportunities

Dynamic and collaborative work environment

Job summary

A dynamic technology startup is seeking a hands-on Full-Stack Engineer to build and manage applications end-to-end. Responsibilities include feature development across the stack, managing infrastructure and deployment, and collaborating with founders on product direction. Candidates should have strong skills in TypeScript, Python, and experience in data engineering. This role offers a competitive compensation package with significant equity opportunities.

Qualifications

• Experience building full-stack applications from scratch.
• Proven skills in data engineering and orchestration.
• Comfortable with ambiguity and high ownership in startup environments.

Responsibilities

• Design and implement full-stack features.
• Manage infrastructure and deployment pipelines.
• Collaborate closely with founders on product direction.

Skills

Strong TypeScript

Python

Next.js

DevOps

CI/CD

Data engineering

Monorepo

Agentic workloads

Tools

Docker

Kubernetes

Airbyte

Fivetran

dbt

Dagster

GraphQL

REST

Job description

Compensation: $125k+ base & meaningful founder-level equity

Overview

We launched a few weeks ago and early feedback/demand has been strong. We’re backed by founders from Apollo, HubSpot, Typeform, and more and are looking for one killer engineer to help us build something people truly want & need. We need a hands-on, full-stack builder who can ship end to end, own deployments, and stand up data and agent infrastructure that scales from the first customers to the next hundred.

What you’ll do

• Build end to end: Design, ship, and own features across the stack (frontend, backend, data, agents).
• Own infra & deployment: Stand up CI/CD, environments, and runtime reliability. You will be the person who gets code to prod and keeps it healthy.
• Monorepo stewardship: Set conventions, tooling, and module boundaries for a fast, clean monorepo.
• Data pipelines: Design and run ETL/ELT flows that are observable, testable, and cost-aware.
• Agentic systems: Build and ship agent workflows, tools, and evaluators with clear success metrics.
• Quality & speed: Write design docs, add tests and telemetry, and move quickly without breaking trust.
• Founder collaboration: Work directly with founders on product direction and customer feedback loops.

You might be a great fit if you have

• Full-stack chops: Strong TypeScript and Python. You’ve shipped serious Next.js frontends and service backends (REST or GraphQL).
• Data engineering: Practical ETL/ELT design, orchestration, schema versioning, and monitoring. Familiar with tools like Airbyte/Fivetran, dbt, Dagster/Airflow (or equivalent approaches).
• DevOps experience: CI/CD (GitHub Actions or similar), containers, infrastructure as code, and either Kubernetes, ECS, or a PaaS. Comfortable owning prod.
• Monorepo experience: Turborepo or Nx, shared packages, caching, build optimization.
• Agentic workloads: You’ve built and deployed LLM agents or tool-using workflows. You understand prompts, tools, memory, and failure modes.
• Agent evals: You’ve set up evaluation harnesses and regression suites for agents or LLM features (e.g., offline eval sets, trace scoring, task success metrics).
• Startup context: 0→1 or early-stage experience at Seed to Series B companies. You are comfortable with ambiguity, high ownership, and direct customer exposure.

Other great experience

• Mastra experience (or similar agent frameworks) and having “fielded” agents in production.
• Porter (Porter.run) or comparable PaaS experience for app delivery.
• Observability for LLMs: Langfuse, LangSmith, OpenTelemetry, or custom tracing.
• RAG and data systems: Vector stores, embeddings, chunking strategies, and retrieval evals.
• Security & auth: SSO/OIDC, secrets management, role-based access, audit trails.

Our stack (indicative)

• Frontend: Next.js, React, TypeScript
• Backend: Node/TypeScript and Python services, REST/GraphQL
• Data: Postgres, Redis, warehouse/lake, ETL/ELT with orchestration
• Infra: Docker, CI/CD with GitHub Actions, IaC, Kubernetes/ECS or PaaS (e.g., Porter/Vercel)
• AI/Agents: Mastra or equivalent, OpenAI/Anthropic, evaluation harnesses, tracing/observability

How we work

• Bias to action, tight feedback loops with users, written design docs, and pragmatic testing.
• Clear ownership, transparent metrics, and direct access to founders.

We like to work hard, have fun, and take care of each other.

Note: This is a refined job description for clarity and formatting; the original notes about general postings have been removed to focus on the role.