<aside> 💡

Why this page exists. Before you touch the repo, spend 15 minutes here. Task 1 explicitly asks you to read the strategy first, the code second and to tie every infrastructure initiative back to a business goal. This is the condensed business context so you can do exactly that.

</aside>

🏥 What Parallel does (in one minute)

Parallel builds a network of AI agents that operate inside hospital systems to handle administrative work no human should still be doing by hand. We start where trust and financial value are highest — medical coding (turning a patient's hospital stay into the billing codes that determine hospital revenue) — and expand horizontally across the hospital ecosystem.

The key technical bet: instead of building heavy integrations, our agents use existing hospital software directly (computer use + LLMs), so we add value on top of the legacy systems hospitals already run, with near-zero integration friction.

🧱 Why infrastructure is the business

Parallel's edge is near-zero marginal cost per new client / per new agent. That single sentence should drive how you read the whole case:

• Every new hospital client should be onboarded and deployed with as little bespoke, manual work as possible.
• Reliability has to hold as we scale to 100+ clients, not just for a handful of pilots.
• We handle French hospital data, so security, data residency, and auditability are non-negotiable, not nice-to-haves.

<aside> 🎯

So the roadmap the team values ranks scalable, repeatable, safe client onboarding/deployment and reliability at scale high — and treats gold-plating / premature sophistication as a distraction.

</aside>

📈 Where the business is right now (2026)

We're on a steep run-rate trajectory. Revenue is the metric being optimized hardest right now, and infra choices should protect that trajectory and existing commitments.

<aside> 🗓️

The case assumes "now" = start of Q3 2026. So the commitments most relevant to your next-quarter roadmap are the EO Q3 → EO Q4 ones: MCO scaled to 100 clients, Screenflow write-back into hospital systems, SMR going live, and the sovereign-cloud move.

</aside>

🗺️ The 2026–2028 arc (roadmap, lightly)

• H1 2026 — Real-time data access: Screenflow live on first clients; primo-codage live; assistant launched.
• H2 2026 — Coding expansion in France: all-in-one primo + opti on MCO and SMR; scale Screenflow to most clients and beyond document retrieval.
• H1 2027 — Platform expansion: billing automation agent; US expansion; better agent monitoring & orchestration.
• H2 2027 — Agent orchestration: orchestrate agents with custom triggers & scenarios.
• 2028 — Custom agents and a successful US expansion.