Healthcare • Fractional CTO

Rupa Health scaled from $100K to $5M ARR during Seed to Series A

HyperNest Labs stepped in as fractional CTO during the early stages to harden infrastructure, lead engineering hiring, and build the technical foundation that supported Rupa Health's rapid growth.

$100K → $5M

ARR scale

300%+

Traffic absorbed

<0.5s

Core web vitals

4 → 15

Engineering team growth

Book a fractional CTO consult

Client snapshot

Company
Rupa Health
Industry
Healthcare SaaS
Stage
Seed → Series A
Headquarters
San Francisco, CA

Timeline

  1. Week 1

    Stabilized infra, set up observability, and triaged debt.

  2. Month 2

    Led hiring loops for full-stack engineers and implemented new sprint rituals.

  3. Month 4

    Completed technical diligence materials for Series A investors.

  4. Month 6+

    Continued scaling infrastructure and team as ARR grew to $5M.

Problem → Action → Outcome

A quick operator-level summary of what changed.

Problem

A fast-growing healthcare marketplace was experiencing 300% traffic spikes without a dedicated CTO, clear architecture ownership, or investor-ready documentation during the critical Seed to Series A phase.

Action

HyperNest embedded as a fractional CTO during the early stages, re-architecting key services, installing observability, and building the hiring + diligence playbook alongside the founders.

Outcome

Rupa Health scaled ARR from $100K to $5M, rode out the surge with 99.99% uptime, and built the technical foundation that supported future growth. The company was later acquired by Fullscript.

The challenge

Demand for Rupa Health’s lab marketplace exploded almost overnight. The company had found genuine product-market fit in root-cause medicine, connecting functional-health practitioners with specialty lab tests through a single ordering interface. But the engineering team that had built the initial product was now drowning in incidents. Every week brought a new fire: API timeouts during peak ordering windows, database connection pool exhaustion under moderate load, and a growing backlog of partner integration requests that nobody had bandwidth to address. The monolithic Rails application that powered the platform had been written for speed-to-market, not for the kind of traffic the business was now attracting. The stakes were unusually high for a Seed-stage company. Rupa Health was preparing for its Series A raise, and prospective investors were asking pointed questions about infrastructure resilience, engineering velocity, and technical debt. The founding team had no architecture documentation to share, no formalized SLOs, and no observability beyond basic server logs. When investors asked whether the platform could handle a 3x traffic spike during Q1 lab-ordering season, the honest answer was that nobody knew. Without a credible technical narrative, the fundraise risked stalling or being discounted. Compounding the infrastructure risk was a hiring bottleneck. The team had grown from two engineers to four, but the company needed closer to fifteen to execute on the product roadmap for Lab Ordering, Telehealth integration, and expanded test catalog coverage. There was no structured interview process, no leveling framework, and no engineering manager to run hiring loops. Senior candidates were dropping out mid-process because the interview experience felt ad-hoc. Every month of delayed hiring meant features slipping, which meant practitioners choosing competitor platforms that could onboard faster.

  • Traffic tripled during seasonal Q1 lab-ordering spikes with limited autoscaling, causing API timeouts that directly blocked practitioner orders
  • Series A investors requested architecture documentation, incident postmortems, and SLO dashboards that did not exist in any form
  • High-touch onboarding for new lab partners required manual configuration, creating a six-week backlog that was growing weekly
  • No dedicated technical leader to manage hiring pipelines, define leveling frameworks, or enforce code quality standards across the growing team
  • Database connection pool exhaustion was occurring under moderate load, pointing to deeper architectural problems in the monolithic Rails application
  • Engineering morale was declining as the team spent more time firefighting than shipping features, with on-call rotations effectively meaning everyone was always on call

What we built

We embedded as fractional CTO in the first week with a clear mandate: stabilize the platform, build the team, and produce investor-ready technical documentation within four months. The engagement started with a 72-hour architecture deep dive. We mapped every service dependency, identified the top ten reliability risks ranked by blast radius, and drafted an incident response plan that the team could execute immediately. The single highest-impact change in the first sprint was implementing connection pooling via PgBouncer and adding request-level timeouts to the API gateway, which eliminated the most frequent class of production incidents within days. From there we reorganized the engineering team into domain-aligned pods: Lab Ordering, Practitioner Experience, and Platform Infrastructure. Each pod got a clear charter, a Datadog dashboard, and a weekly demo cadence. We chose Datadog over alternatives like Grafana or New Relic because Rupa's AWS-heavy infrastructure meant Datadog's native integrations would give us deep visibility without custom instrumentation. We defined SLOs for each pod's core user journey (e.g., lab order submission p99 latency under 800ms, practitioner search under 200ms) and wired PagerDuty alerts to fire only when SLO burn rates indicated real user impact rather than noise. Architecture Decision Records (ADRs) became mandatory for any change touching shared services, giving the team a written record of technical trade-offs that investors could review during diligence. Hiring was treated as an engineering problem, not an HR afterthought. We built a structured interview loop with a take-home exercise calibrated to two hours, a system design round, and a values-fit conversation with the founders. We defined three engineering levels with clear compensation bands so that offers could go out within 48 hours of a final round. To source candidates, we partnered with two boutique recruiters who specialized in healthcare SaaS and ran a referral program that paid out within 30 days. The combination of a faster process, clearer leveling, and a compelling technical story (including the ADRs and SLO dashboards) cut time-to-hire from 90 days to 35. Over four months, the team grew from four engineers to fifteen, with every hire passing a rigorous bar that the founders and existing team had agreed on in advance.

Introduced Architecture Decision Records (ADRs) as a mandatory practice for shared-service changes, creating an audit trail that investors reviewed during diligence
Refactored the monolithic Rails application into modular domain services with shared libraries, reducing deployment risk and enabling independent pod-level releases
Implemented Datadog dashboards with custom SLOs for every core user journey, plus PagerDuty alerting tuned to SLO burn rates rather than raw error counts
Designed and ran structured hiring loops with take-home exercises, system design rounds, and 48-hour offer turnaround for frontend, backend, and data engineering roles
Created board-ready weekly engineering scorecards and investor diligence materials covering architecture, reliability, security posture, and hiring velocity
Deployed PgBouncer for connection pooling and added API gateway timeouts, eliminating the most common class of production incidents within the first sprint
Established performance budgets for Core Web Vitals (LCP under 500ms, CLS under 0.05) enforced in CI, ensuring the practitioner portal stayed fast as features shipped

Impact

Rupa Health continued shipping new revenue lines including Lab Ordering, Telehealth integration, and expanded test catalog coverage while maintaining 99.99% uptime across the entire growth period from $100K to $5M ARR. The platform absorbed a 300% traffic surge during Q1 lab-ordering season without a single customer-facing incident, a direct result of the autoscaling policies, connection pooling, and SLO-driven alerting we put in place. Before the engagement, the team averaged two severity-1 incidents per month. After the infrastructure stabilization, that number dropped to zero for six consecutive months. Engineering morale visibly improved: voluntary attrition went from two departures in a quarter to zero, and internal surveys showed that engineers felt they were shipping features rather than fighting fires. The Series A fundraise closed with strong investor confidence in the technical foundation. Board members and lead investors received weekly engineering scorecards that tracked deployment frequency, incident counts, SLO adherence, and hiring pipeline conversion rates. These scorecards became a template that Rupa continued using long after our engagement ended. Time-to-hire for senior engineers dropped from 90 days to 35 days, and the team grew from four to fifteen with a consistently high hiring bar. The technical documentation, ADR library, and architecture diagrams we produced became core artifacts in subsequent fundraising rounds. Rupa Health went on to become a market leader in functional medicine lab ordering, and the company was later acquired by Fullscript, a publicly verifiable event that validated the durability of the technical and organizational foundation we helped build during those early critical months.

ARR grew from $100K to $5M with zero severity-1 incidents during the entire growth surge, compared to two per month before the engagement
Board and Series A investors received weekly engineering scorecards tracking deployment frequency, SLO adherence, and hiring velocity
Time-to-hire for senior engineers dropped from 90 days to 35 days through structured loops, clear leveling, and 48-hour offer turnaround
Engineering team grew from 4 to 15 while maintaining a consistently high hiring bar agreed upon by founders and existing team
Platform absorbed 300% seasonal traffic spike with zero customer-facing incidents after autoscaling and connection pooling improvements
Rupa Health was later acquired by Fullscript, validating the long-term durability of the technical foundation built during the Seed to Series A phase

Stack & capabilities

Tools, platforms, and competencies we owned for this engagement.

Frontend

  • Next.js
  • React
  • Tailwind

Backend

  • Node.js
  • GraphQL
  • PostgreSQL

Infra & Ops

  • AWS
  • Terraform
  • Datadog
  • PagerDuty

Aravind is in the top 1% of engineers I've hired. He supported us from $100K to $5M ARR, keeping us shipping through crazy growth and investor scrutiny.

Tara Viswanathan

Co-founder & CEO, Rupa Health

Need these outcomes for your startup?

We'll replicate the playbook and customize it to your goals.

How this applies to other startups

  • If you are an early-stage healthcare or SaaS startup facing sudden demand, you can borrow a pre-Series A CTO playbook instead of learning scaling lessons the hard way.
  • Founders preparing for a major fundraise can use the same architecture, metrics, and diligence patterns we deployed at Rupa Health during their Seed to Series A journey.
  • Teams without a full-time CTO can still run board-ready engineering and reliability rituals by embedding fractional leadership plus a small IC bench.
  • If your monolithic application is hitting connection pool limits under moderate load, the PgBouncer and API gateway timeout pattern we deployed can buy you months of runway before a full re-architecture.
  • Engineering teams struggling to hire senior talent can adopt the structured interview loop, leveling framework, and 48-hour offer turnaround process that cut Rupa's time-to-hire by more than half.
  • Any startup entering investor diligence can use the weekly engineering scorecard template (deployment frequency, incident count, SLO adherence, hiring pipeline conversion) to proactively answer the questions VCs will ask.