AI Analytics Platform for Customer Service
An analytics platform for customer service and support teams, with views for agents, supervisors and leaders, and AI that explains what changed and why in plain language.
- Client
- serviceMob
- Industry
- Customer Service Analytics
- Timeline
- 2026 to present
- Platforms
- Web
Who uses it
- Agents
- Supervisors
- Operations leaders
The brief
What serviceMob needed
serviceMob helps customer service and support organizations model the customer experience as data, so they can prevent demand instead of only handling it. It measures whole experiences rather than single contacts, with metrics like minutes per resolved experience and contacts per resolved experience.
serviceMob had a prototype of its analytics product and needed an engineering team to turn it into a production application for agents, supervisors and operations leaders, including the AI features the prototype showed. serviceMob kept ownership of its data engineering, analytics and models.
VantaSoft worked as serviceMob's engineering team, with a fractional CTO responsible for architecture, delivery and technical risk.
What we built
What we built for serviceMob
-
Supervisor dashboards
Team averages and trends for resolution time, first-contact resolution, transfer rate, handle time, CSAT and NPS, with team rankings, rank changes over time and the top issue categories.
-
Agent dashboards
Each agent's metrics and ranking, comparison with peers, the issue types they resolve least often, and their CSAT and NPS trends.
-
Targets and weighting
Metric targets by team, business unit and skill group, metric weights with validation, and previews with histograms, trend overlays and target bands. Every change is audited.
-
Reports and prioritization
Effort against volume to rank what to fix first, contacts per resolved experience, issue-level rollups with sentiment and CSAT, and an enterprise overview.
-
mobAI
Leaders ask questions of their data in plain language and drill down to the issue type, the team and the agent.
-
Run Highlights and Data Journals
A written account of what changed this period, where and why, for each business unit, in place of hunting through dashboards.
-
AI agent architecture
AI modules built on the Claude Agent SDK, reading call transcripts and aggregated call data through MCP servers, with each customer's settings and guardrails applied per session.
-
Access and security
Role-based access control and security practices aligned with SOC 2 Type II expectations.
-
Foundations for serviceMob's team
API contracts, engineering standards, code review and documentation, so serviceMob's own engineers can own and extend the platform.
Results
What it changed
-
Prototype to production
serviceMob's prototype became a production application for agents, supervisors and operations leaders, AI features included.
-
AI that explains the numbers
mobAI and Run Highlights tell leaders what changed and why, and let them ask follow-up questions, instead of leaving them to read dashboards.
-
Built for serviceMob to own
Every line of code belongs to serviceMob, with the standards and documentation its team needs to extend the platform.
-
Secure by design
Role-based access and security practices aligned with SOC 2 Type II, from the first phase.
Where it stands In production. VantaSoft continues to support and maintain the platform, with a fractional CTO.
Stack
- Node
- Express
- Claude Agent SDK
- MCP
- ClickHouse
- AWS Athena
What we did
- Engineering Team as a Service
- Fractional CTO
- LLM Integration
You run the business.We run the tech.
A long-term partner, not a one-off project. The IP is always yours. It starts with a conversation.