A scientific safety software platform for research institutions came to Armakuni for aI-led platform modernization: containerized infrastructure, GenAI chemical profile automation, and WCAG accessibility compliance.
The challenge
A scientific safety software platform for research institutions came to Armakuni after completing a Rackspace to AWS migration, needing a partner to take the platform from migrated to modernized. A containerization PoC expanded into a 4-SOW engagement spanning infrastructure, GenAI strategy, feature quality, and accessibility compliance, with Armakuni now operating as A scientific safety software platform for research institutions's embedded engineering partner across every modernization track.
Scale: Two platforms serving research customers across the U.S., spanning a legacy Drupal and PHP stack and a modern Vue.js stack.
What we built
Armakuni delivered aI-led platform modernization: containerized infrastructure, GenAI chemical profile automation, and WCAG accessibility compliance. The engagement ran as multiple fixed-scope SOWs: Containerization PoC, GenAI Use Case Discovery Assessment, Feature Audit, and Accessibility Remediation Pod. Accessibility runs on an embedded pod model (3 full-stack engineers, 1 QA engineer). AWS funded the work at $19K.
The outcomes
6 measurable outcomes shipped across the engagement. The ones that moved the business the most:
Containerized architecture was validated on Amazon EKS and Amazon Aurora MySQL. Every new customer now onboards on automated, scaled infrastructure instead of the hand-configured VMs that had capped growth, with database operations a commodity the platform absorbs.
A 5-to-60-minute manual chemical profile task became a funded GenAI implementation. Evaluated foundation models, a compliance-by-design architecture, and an executive-approved business case replace a permanent ceiling on catalog growth, with the SOW starting from validated choices.
A prioritized remediation backlog was delivered in two weeks. Against a bug backlog accumulating across Jira and Zendesk, engineering shifted from chasing tickets to a sequenced plan of root-cause fixes, each with rollback criteria so the work is durable.
Quick-win mitigations during the feature audit reduced customer pain within hours. Configuration changes, guardrails, and feature-flag adjustments gave customers relief while the long-term fixes were being scoped rather than waiting on them.
27 accessibility defects are on track for remediation across two platforms in parallel. An embedded pod works inside the engineering cadence so WCAG 2.1 compliance lands ahead of the regulatory deadline, leaving the internal team stronger on accessibility than it started.
One AWS Premier Partner is embedded across every initiative. Infrastructure modernization, AI strategy, engineering quality, and accessibility now run through a single delivery team, a relationship grown from one containerization PoC into a sustained partnership.
Built on AWS
The production environment runs on 8 first-party AWS services. Delivered under our AWS Migration and Modernization, AI Services, DevOps Services competencies.
What's next
GenAI chemical profile tool implementation on Amazon Bedrock. Continued pod engagement for ongoing modernization work.