Senior / Lead Backend Engineer

Company:

Location:

Type:

Compensation:

Engineer: $2000USD/month + 1% equity

Upon promotion to CTO: $4000USD/month + 2% equity

booked appointments

scale‑ready backend platform

• Multi‑tenant architecture

  with tenant isolation, role‑based access (RBAC), and per‑clinic secrets (env‑less runtime or AWS Secrets Manager).

• Onboarding automation

  that reduces per‑clinic setup to

  ≤60 minutes

  : provider/location/appointment‑type sync, ID mapping, test calls, and health checks.

• Hardened tool endpoints

  used by the voice agent (Retell function calling):
• availability_search, appointment_book, appointment_reschedule, appointment_cancel,
• patient_find_or_create, note_create, warm_transfer.

• Reliability controls

  : idempotency keys, timeouts, retries with backoff, circuit breakers; graceful fallbacks + warm transfer.

• Observability & SLOs

  : structured logs, metrics, tracing; dashboards for

  p50/p95 latency

  , error rates, booking success %, transfers, cost per minute/call; alerts to Slack.

• Security & compliance

  : PHI minimization, at‑rest and in‑transit encryption, access logging, data‑retention policy, BAA‑aware configuration.

• Cost guardrails

  : per‑tenant budget meters for voice minutes/LLM/TTS usage and anomaly alerts.

KPIs you’ll move:

• Median tool‑call latency

  < 800 ms

  (p95 < 1500 ms)

• ≥ 80%

  booking/reschedule success without human handoff (eligible calls)

• 99.9%+

  middleware availability

• < 1%

  tool‑level error rate (after retries)

• ≤ 60 min

  time‑to‑onboard a new clinic (target

  30 min

  by week 6)

• Design, implement, and document

  multi‑tenant

  REST/JSON services consumed by the voice agent.
• Integrate

  NexHealth

  now; design extension points for

  direct PMS

  (OpenDental/Dentrix/Eaglesoft/Dolphin) later.
• Build

  sync jobs

  to keep providers/locations/appointment types up‑to‑date (with caching via Redis, invalidation, and backfills).
• Implement

  idempotent booking flows

  with conflict detection and safe retries; log every state transition.
• Stand up

  observability

  (metrics/logs/traces) and alerting; define

  SLOs/SLA

  and on‑call basics.
• Ship

  CI/CD

  with linting, tests (unit, contract, integration), and minimal load tests.
• Enforce

  secrets management

  , least‑privilege IAM, and a clean

  audit trail

  .
• Partner with our conversation designer to refine tool schemas and edge‑case flows (insurance screening, multi‑location routing).
• Mentor a mid‑level engineer and coordinate with ops for smooth rollouts.

• 5–8+ years

  building production backend systems (you’ve owned a system in prod).
• Expert in

  Node.js (TypeScript)

  or

  Python

  (FastAPI/Nest/Express).
• Deep experience with

  external API integrations

  (auth, pagination, rate limits, webhooks).

• Postgres

  (schema design, migrations) and

  Redis

  (caching, locks).
• Production reliability patterns: retries/backoff, timeouts,

  idempotency

  , circuit breakers.
• Observability: metrics, tracing, log correlation; incident triage.
• Security/compliance mindset; comfortable handling sensitive data flows.
• Strong written English; crisp architectural docs and PRs.

• Retell AI (or similar voice/LLM with function calling and barge‑in),

  Twilio/SIP

  .
• NexHealth or other healthcare scheduling APIs; PMS/EHR familiarity.
• HIPAA/PIPEDA exposure, SOC 2‑style controls.
• OpenTelemetry, Prometheus/Grafana, Sentry; AWS/GCP; Terraform; Docker/Kubernetes.
• High‑volume, low‑latency systems experience.

• Runtime:

  Node.js (TypeScript)

  or

  Python (FastAPI)

• Data:

  Postgres, Redis

• Infra:

  AWS (ECS/EKS or Fargate), Terraform, GitHub Actions

• Integrations:

  Retell AI (voice), NexHealth (scheduling), Twilio/SIP (telephony)

• Observability:

  OpenTelemetry + Prometheus/Grafana or Cloud provider equivalents

• Remote‑first; async‑friendly;

  4+ hours

  overlap with Pacific time.
• Code in company repos,

  NDAs/PIAs/BAAs

  , DCO/CLA, and strict access hygiene.
• We optimize for reliability and patient privacy over quick hacks.

1. Intro (20–30 min):

   Your background, past scale/reliability wins.

2. Take‑home (90 min, paid for finalists):

• Implement availability_search + appointment_book against a stubbed NexHealth‑like API.
• Include idempotency keys, retries with backoff, timeouts, and basic tests.
• Provide a short

  runbook

  and a

  dashboard sketch

  for p95 latency & error‑rate alerts.

1. Deep‑dive (60 min):

   Review your code; discuss multi‑tenant design, secrets, SLOs, and cost control.

2. Final (30–45 min):

   Collaboration & comms.

info@apexdentalsystems.com

• CV + GitHub/portfolio
• 5–10 lines on a system you made

  multi‑tenant

  (what changed?)
• A time you prevented

  double bookings

  or handled idempotency at scale
• Your preferred stack (Node+TS or Python), availability, and comp expectations