Case studies
Proof, not positioning.
Every engagement below is a production system we shipped, ran, or handed over — with the constraints we worked inside and the figures we measured afterwards. Read them for the shape of the work, not the logo wall.
Engagements
10
Industries
29
Service lines
59
Measured outcomes
50
Outcome library
Browse the work by the results it created.
Each story highlights the platform pressure, the engagement model, and the measurable changes that followed.
Undisclosed Specialty Insurer
Insurance
Candidate use cases assessed
19 → 1 funded
6 weeks
Nineteen AI Ideas In, One Funded Build Out — In Six Weeks
A six-week discovery engagement scored 19 candidate AI use cases against evidence — data, workflow cost, and regulatory exposure — and sent exactly one of them to a funded build.
Undisclosed Industrial Distributor
Industrial Distribution
Orders entered without human touch
0% → 68%
week 14
68% of Purchase Orders Now Enter the ERP Untouched
A 14-week implementation put an order-entry agent into production that reads inbound purchase orders, validates them against the ERP, and writes 68% of them straight through.
Undisclosed Professional Services Firm
Professional Services
Weekly active AI users
9% → 71%
week 16
From 9% to 71% Weekly AI Use Across 217 People
A 16-week company-wide enablement program took a professional services firm from 9% to 71% weekly AI use by teaching each job family on its own work and leaving 14 trained champions behind.
Cosellr
Rapid MvP
Median voice response latency (end-to-end)
0.9–1.8s
steady-state target
Real-time voice agents that turn conversations into pipeline
Cosellr’s voice agents use Pipecat and Daily.co to run real-time sales calls with tool-backed reasoning, guardrails, and production-grade observability.
Cosellr
Rapid MvP
Time to MVP (net-new build)
8 weeks
end-to-end delivery
From zero to AI-first sales platform in 8 weeks
We built Cosellr’s AI-first sales platform from scratch in 8 weeks—launching a production-ready MVP across prospecting, outreach, nurturing, and in-call agent assistance.
Trellis Connect / Trellis Technologies
Insurtech
Annualized cloud cost savings
$3.2M+
Slashed 26% off Annual Cloud Costs in 4 Weeks
Epsilon helped Trellis Connect eliminate $193k in annual cloud waste in just four weeks by fixing the engineering systems that caused costs to creep.
RentPath
Cloud Native Development
Applications migrated
30
30 days
Rapid Kubernetes Migration from Docker Swarm
RentPath moved 30 apps from Docker Swarm to Kubernetes in 30 days with zero downtime.
BetterCloud
Platform Engineering
Infrastructure cost reduction
17%
post-migration
From Mesos & Marathon to Kubernetes
BetterCloud modernized its platform by migrating from Mesos/Marathon to Kubernetes while adopting GitOps and production-grade observability.
Ramboll SHAIR
Platform Engineering
Monthly infrastructure savings
58%
within 60 days
Infrastructure-as-Code with Terraform
Ramboll SHAIR cut cloud costs dramatically by moving to Terraform-driven, ephemeral environments with automated provisioning and teardown.
Trellis Connect
Legacy Modernization
Cold start time
10 min → 45 sec
post-optimization
Mobile Emulation and Secure Data Collection in Kubernetes
Trellis Connect achieved real-time readiness by cutting cold starts from ~10 minutes to ~45 seconds on Kubernetes.
Bring us the AI initiative that needs to become real
Let’s build AI your organization can trust, operate, and scale.
Share the workflow, product, or internal capability you want AI to improve. We will help you identify the right path, build it responsibly, and leave your team capable of operating and expanding it.
AI readiness review
Evaluate your opportunities, workflows, data, architecture, security posture, and organizational readiness.
Production AI sprint
Design, build, evaluate, and deploy a focused AI system with production guardrails from the beginning.
Embedded AI partnership
Add senior engineering capacity to accelerate your roadmap, establish internal standards, and scale adoption safely.