CHI runs the full stack of a live event: ticketing, cashless payments, POS, wristbands, vendor settlements, coupons, CRM, and analytics — for festivals, venues, and large-scale events.
When our platform has a bad day, thousands of people can't buy a drink. That's the level of “production” we mean.
// A small team, 100+ AI agents
A small team,
100+ AI agents
CHI is built as an autonomous company: a small human team directing a workforce of AI agents that run every repeatable process — code, review, release, QA, on-call, support, reconciliation, fraud review, outreach. Humans set direction and taste; agents execute. Already proven internally — a much smaller team shipping several times faster than the org it replaced.
We're hiring a human operator for that agent workforce — an AI-native engineer (dev + DevOps). Not someone who “uses AI sometimes,” but someone who builds, optimizes, and directs AI agents and their loops to ship software as their primary way of working: product features, coding, testing, debugging, infrastructure, incident analysis, and documentation, all agent-driven. You'll be one of a handful of humans in the company, work directly with the founders, and own real systems — and the agents that run them — end to end.
// What this job actually looks like
A real week here, not a hypothetical one
Ship an organizer-facing feature end to end — Postgres data model, NestJS API, React backstage UI, tests, release notes — driving AI agents through every layer and owning how it feels.
Design a new ticketing or payments flow that thousands of attendees will actually use — from spec to shipped in the app and on the POS — then watch it live.
Trace a failed card payment through the gateway, backend logs, and the provider's dashboard — then ship the fix and a reconciliation script the same day.
Take a change from PR to production through our versioned pipeline: conventional commit → automated release → OCI artifact → test → staging → production promotion.
Build an AI agent or CI workflow that reviews PRs, runs QA, or drafts release notes and incident reports — so the team never does that manually again.
Respond to a production incident — a disk-full VM, a missed webhook — fix it, backfill the data, and automate the check that catches it next time.
If that list makes you want to open a terminal, keep reading.
// Our stack (the real one)
We list this because we want people who like this work
Backend
Node.js / TypeScript (NestJS), Bun
Frontend
React (organizer backstage + web apps), PWA and native mobile
Data
PostgreSQL (WAL + PITR), Redis, Kafka, Cassandra, YugabyteDB edge replication
Infrastructure
Docker Compose on dedicated Hetzner VMs. Zero-trust networking via Tailscale, hardened host firewalls.
Gateway & Identity
KrakenD API gateway, Keycloak, HashiCorp Vault (agent templates, AppRole auth)
Payments
Stripe Connect, Hyperswitch (PCI vault + cloud KMS), Xendit, Stripe Terminal / tap-to-pay. Real money, real settlements.
CI / CD
GitHub Actions (60+ workflows), release-please, OCI config artifacts, gated test → staging → prod promotion
Observability & Security
Grafana, Loki, Prometheus, Promtail, Wazuh HIDS, ClamAV — operating inside PCI DSS scope
AI Tooling
Custom coding agents, multi-agent workflows, MCP servers, Paperclip, Hermes agents, LiteLLM proxy, local / self-hosted LLMs
You don't need experience with all of it. You need to be the kind of engineer who picks up any of it fast — with AI doing the heavy lifting.
// The operator bar
Full-stack skill plus business judgment — applied by orchestrating agents
This is the core of the role. At CHI, “operator” means orchestrating agents instead of doing everything by hand. We expect that you already:
Ship production code daily with AI coding agents — as your default mode, not a party trick
Design and run multi-agent workflows: review, QA, deployment watching, incident drafting — improving from every run
Use AI to debug incidents: feed it logs, configs, stack traces, and drive it to root cause
Build automation around AI, not just with it — agents, review bots, report generators, internal tools
Know where AI is wrong: you verify, test, and own the output. Agent speed with human judgment.
Write documentation and runbooks as you go, because generating them costs you minutes
Your job is not to cover a vertical by hand — it's to design, ship, and manage the agent workforce that runs it, and step in with judgment where only a human can.
// What we expect
From you
Strong backend skills (TypeScript / Node.js or equivalent — willing to go deep in ours)
Real DevOps: Docker, Linux, networking, CI/CD, logs, monitoring — and the scar tissue of debugging prod at 1 a.m.
Product sense: you can take a feature from spec to shipped UI and API, and you care whether it actually feels right for organizers and attendees
Solid grasp of APIs, databases, queues, authentication, and distributed systems
Ownership: you see a problem, fix it, automate it, document it — into the shared knowledge base and the agents' self-learning loops
Business judgment: operators own outcomes across product, ops, and finance — not just code
Clear written English — most coordination is async and written
Care for stability and security: we handle payments; sloppiness is expensive
// Nice to have
Bonus signal
// Why CHI
Leverage that a normal team can't offer
Real stakes
Live events, live payments, live vendors. Your work runs in front of crowds, not in a sprint backlog.
Agents, not seats
Few humans, each given a workforce of agents. Your leverage is a multiple of a normal team's.
Full ownership
No ticket-shuffling, no three layers of approval. Product, backend, infra, and AI tooling in the same week.
The autonomous-company frontier
Most companies write think-pieces about AI-native teams. We are one — and you'll help define how it operates.
Skin in the game
Revenue share on top of base — when the platform earns, you earn. Each person's upside should look like it.
Bali base, remote functions
We keep an Indonesia/Bali office base for live collaboration, and support remote work where timezone overlap, ownership, and customer responsibility stay strong.
Event access
Our platform runs festivals. You'll see your work in the field.
Tooling budget
We pay for the AI tools and hardware that make you fast — including local-inference rigs alongside frontier APIs.
// How to apply
Your CV or LinkedIn
GitHub or examples of your work
A short note on how you use AI in your daily workflow — concrete: which tools, for what, and one thing you built or automated with them
One example of infrastructure, automation, or DevOps work you own end to end
We read the AI-workflow note first. A great one gets a reply within days — “I use ChatGPT to write emails” does not.