Founding Full Stack Engineer

Full Stack Developer (Founding Engineer)

AnoNa Labs · Full-time · Remote (India)

AnoNa Labs is registered and headquartered in San Francisco. The company is remote-first; this role is based in India.

WHAT WE ARE BUILDING

AnoNa Labs is building the company brain: everything your organization learns, remembered forever, working for every person in it. One stack, two ends, and we own both.

• AnoNa Memory is the brain. In Private Beta today. A memory backend for AI agents: it ingests what an organization produces, extracts facts and entities, detects contradictions, and supersedes stale information. Any agent or application can store and recall through its APIs, SDKs, and MCP servers, across personal, team, org, and project memory.
• The client apps are the body. A desktop app (macOS, Windows) and a companion web app that work alongside knowledge workers all day: they connect to the tools people already use (Slack, email, calendar, meetings, documents), compound everything into live personal and team context on AnoNa Memory, and put that context to work through agents that brief, draft, research, and act.

The two ends make each other stronger: every seat of the app exercises the memory at production scale, and the memory makes the app something no thin wrapper can copy. This is the founding engineering role. You build the client-app surface end to end and carry a real share of the backend.

WHAT YOU WILL OWN

• The client apps, end to end. The Electron desktop app (macOS and Windows) and the web app, on a shared core. Main/renderer design with secure IPC, context isolation, sandboxed renderers. The full release pipeline: build, sign, notarize, auto-update with staged rollouts and rollback. Crash reporting, diagnostics, and performance budgets you set and enforce (cold start, idle memory, battery). The app runs all day beside the user; it has to feel invisible.
• Agents and orchestration. The layer that turns stored context into work: multi-step agent runs, deep-research workflows, streaming and tool calling against LLM APIs, and the identity and permissions that keep personal, team, and org context separate.
• Secure integrations to Slack, Google Workspace, and the other tools where work happens: OAuth, webhooks, incremental sync, rate limits. Connectors feed live context into AnoNa Memory.
• The backend. Services on AWS (ECS, EC2, RDS) that power both surfaces: ingestion pipelines, agent orchestration, and the APIs the clients ride on. Contribute to AnoNa Memory's APIs, SDKs, and MCP servers as the apps push them further.

Honest split: roughly 60% client apps and connectors, 40% AWS backend and developer surface. If you only want one half, this is the wrong role; if you want both, it is a rare one. (We actually mean full-stack, not 'I can center a div but please don't make me touch the database').

As the first engineer you also set the foundations everyone else inherits: CI/CD (GitHub Actions), repo and review standards, testing and observability, release engineering (signing, update channels, rollouts), security and secrets hygiene, cost-aware tooling choices, documentation that makes engineer #2 productive in days, and a voice in hiring them. You help drive product direction: what you hear from users becomes roadmap.

MUST-HAVE

• 4-7+ years shipping production code with real end-to-end ownership; CS degree (BS/MS). Expert TypeScript/React/Node; strong working Python. (More than 7 years is fine; this is a senior IC role scoped to an early-stage comp band.)
• You have shipped and maintained software users run all day: a desktop app (Electron or native), or a client where you owned long-running-process problems (memory leaks, offline behavior, background sync, auto-update). A link we can download is ideal; a detailed ownership write-up works too.
• Integration engineering against real third-party APIs like Slack and Google: OAuth flows, webhooks, token refresh, rate limits, retries.
• Cloud and delivery: hands-on AWS (ECS, EC2, RDS) or deep GCP/Azure equivalent; PostgreSQL and Redis; CI/CD with GitHub Actions; testing and observability as habit.
• You have shipped at least one real feature on LLM APIs and understand streaming, tool calling, agentic loops, and their failure modes; you reach for evals to prove what you built actually works.
• AI-native. Coding agents are how you work, not something you dabble in: you drive them across planning, coding, testing, and review, and you can show your setup. One interview round is exactly that walkthrough.

Bonus: deep Electron internals (process model, preload patterns, native modules, major-version upgrades in production) · OS-level context work (global shortcuts, notifications, accessibility, audio/screen capture) · MCP servers or clients; agent frameworks · real-time systems (WebSockets, streaming UX) · knowledge graphs or vector retrieval; AWS Bedrock/AgentCore · local-first storage and sync (SQLite, encryption at rest) · open-source contributions.

HOW WE WORK

AI-native by default: coding agents are daily tools, used to the maximum across the development loop; judgment, architecture, and taste stay yours (picture Mr. Jassy, Mr. Pichai, and Mr. Nadella serving as your personal AI aides). Async-first, with a small overlap window with San Francisco (Pacific time) when it matters. Prototype over debate. Small surface, deep quality. Fully remote within India.

Amazon's Leadership Principles are our operating system; they shape how we build, decide, and interview. The ones we lean on hardest: Customer Obsession (start with the user, work backwards) · Ownership (never "not my job") · Invent and Simplify · Learn and Be Curious · Frugality · Bias for Action · Earn Trust · Dive Deep · Think Big · and above all Deliver Results: ship the right thing, with quality, on time, despite setbacks.

COMPENSATION

Early-stage cash (we are bootstrapped, and honest about it) plus meaningful founding equity; the equity is the point of a founding role. Exact numbers, funding status, and runway are on the table in the first call, before you invest a minute of interview time.

One thing we are clear about: this is a full-time, all-in role. We are not looking for contractors, consultants, or anyone splitting time across commitments.

PROCESS

Decision inside 2 weeks of the first call:

• Intro call: the role, the product (including what we are not saying publicly), funding, and comp. You interview us too.
• Take-home: a small, scoped build close to the real work. (We promise it won't ruin your weekend, and no, we aren't secretly tricking you into building our MVP for free.)
• Technical round: deep-dive on your take-home and on something you have shipped: architecture, decisions, scars.
• AI-workflow round: walk us through your coding-agent setup, workflows, and prompting, live.

Apply with a link to something you have built that we can try (or your ownership write-up), plus a short note on the hardest production problem you have solved (please do not write more than 3 paragraphs or 1 page – there is a real human reading it on the other end). You might be a fit if you have shipped software real people run every day and you want to own a product, not a ticket queue.