Software Project Rescue: Notary.io
A software project rescue for Notary.io. Leanware took over an unstable notarization platform, eliminated session-drop failures, and brought engineering spend back under control.
- Geography
- United States
- Year
- 2024
- Stage
- Growth-stage SaaS
- Team
- 1 senior full-stack engineer, 1 QA, 1 product owner (managed team engagement)
- Duration
- Ongoing
The situation
Notary.io reached out in the situation that usually starts a software project rescue: a previous team had left them with a notarization platform that was unstable under load, and an engineering bill that kept climbing without the reliability to show for it. Notary.io runs the kind of platform where reliability is the product. Customers complete legally binding notarization sessions on it, and if a session drops mid-flow, the trust evaporates.
Three problems compounded: session stability under load, infrastructure that did not scale gracefully, and UX gaps in the notarization workflow itself. On top of those came the compliance and certification requirements of serving regulated transactions at scale. The brief was to stabilize the product and get the spend under control at the same time.
What we built
Leanware took over as a managed team and ran the recovery in the sequence a failing software project needs: take over cleanly, stabilize what is breaking, rebuild the infrastructure underneath, and bring the spend in line while doing it.
Taking over with a lean team
A senior full-stack engineer carried the technical lead, paired with a QA engineer and a product owner, working as a dedicated AI engineering team. The team ran AI-augmented from day one, which is what let three people move at the pace the recovery needed without the headcount the previous arrangement had been billing for. That is where the budget came back into line.
Stabilizing the product
The first target was the session-drop class of bug, the single largest reliability gap on the platform. That work eliminated it. Sessions now run uninterrupted through the full notarization flow.
Rebuilding the infrastructure
The platform moved to Kubernetes on Linode. Compute scales against demand, deployment cadence tightened around GitHub Actions, and observability and incident response are built into the platform rather than bolted on. Document handling was optimized for faster retrieval and download, which mattered because every session ends with the user pulling notarized documents and the slow path was painful. The compliance work covered the certification needs of the regulated context. Tech stack: Vue 3 frontend, Node.js and Laravel backend services, Linode infrastructure, Kubernetes, GitHub Actions.
Outcome
-
Platform stabilized: session-drop defects eliminated, with notarization sessions running uninterrupted end to end
-
Engineering spend brought back under control with a lean, AI-augmented team
-
Infrastructure moved to Kubernetes for graceful scaling under load
The rescue closed out the way a software project rescue should: a stable, scalable, compliance-ready notarization SaaS, with engineering spend that is predictable rather than open-ended. The engagement continues as a managed-team relationship covering ongoing product and infrastructure work. The team that stabilized the platform is the team that keeps shipping on it.
Engagement line
Engagement FAQ
What are the signs a software project needs a rescue?
The pattern Notary.io brought to Leanware is typical: engineering spend climbing month over month, reliability falling instead of rising, and a previous team unable to deliver either. If your platform is unstable under load and the bill keeps growing without results to show for it, the project needs intervention rather than more of the same.
Can Leanware take over a software project from another development company?
Yes. Notary.io came to Leanware after a previous team left the platform unstable and over budget. Leanware took over as a managed team, stabilized the product, and the engagement continues today covering ongoing product and infrastructure work.
How do software project rescue services bring engineering spend under control?
In Notary.io's case, with a smaller team working faster. A senior full-stack engineer, a QA engineer, and a product owner ran the recovery AI-augmented from day one, replacing the larger headcount the previous arrangement had been billing for. Spend went from open-ended to predictable.
Do I need a software rescue consultant or a full replacement team?
It depends on whether the problem is advice or execution. Notary.io needed execution: the session-drop bugs, the infrastructure, and the compliance work all had to be built, not just diagnosed, so a managed team that could take ownership of the codebase made more sense than a consultant producing recommendations.
What happens to the infrastructure during a software project rescue?
Often it gets rebuilt while the product keeps running. Notary.io's platform moved to Kubernetes so compute scales against demand, deployments tightened around GitHub Actions, and observability and incident response became part of the platform rather than an afterthought.