a

Case Study

How Let’s Play Sports Uncovered $100K in Annual AI Savings - in a 2-Week AWS AI Assessment

Let’s Play Sports used a two-week AWS AI Assessment to prioritize four agentic AI use cases, building a board-ready roadmap projecting $100k in annual savings against a $50K build investment and a 400% two-year ROI.

Industry: Sports & Recreation   |   AWS AI Assessment · Sprint 1: Modernization to Agents   |   July 2026

Let's Play Sports multi-facility operations managed by AWS AI agents built by AllCode

The challenge

Manual operations, fragmented financials, and an ESOP deadline

Nearly every core workflow at Let’s Play Sports depended heavily on manual effort. Around 10 FTE were dedicated to consolidating financial data from 19 separate QuickBooks accounts into company-wide P&L reports. League scheduling also required repeated manual corrections, contributing to potential revenue losses. More than 8,100 job applications were reviewed manually each year, while customers had no way to book practices online and after-hours inquiries often went to voicemail or competitors.

These challenges became more urgent as LPS worked toward ESOP readiness by March 31, 2027, requiring cleaner and more auditable financial processes. With 90% of customer interactions happening through text-based channels, slow response times also created customer experience and revenue risks. At the same time, competitive pressure from AI-enabled rivals and a planned migration from SQL Server 2014 to AWS RDS created the right opportunity for LPS to modernize its technology, automate key workflows, and improve operational scalability.

The AllCode solution

A quantified, board-ready AI roadmap in two weeks

AllCode applied AWS’s published four-phase assessment methodology — Use Case Discovery, AI Readiness, AWS Fit Validation, and ROI Business Case — in a two-week engagement.

Four agentic use cases were evaluated end-to-end and ranked by business impact, agentic fit, technical feasibility, and AWS platform fit. Each was delivered with a scoped 90-day POC, success criteria, guardrails, and a per-use-case ROI model. Existing ad-hoc ChatGPT workflows were translated into designs for governed, auditable agents with human-in-the-loop gates.

The first POC candidate, applicant screening and interview coordination, was scoped for a 6–8 week path to live, with a four-quarter roadmap covering all four agents.

Applicant screening & interview coordination

Automate applicant screening and interview coordination to reduce resume review from 15 minutes to 2 minutes and target a reduction in time-to-first-interview from 8 days to 3 days.

Financial reporting & P&L consolidation

Consolidate 19 separate QuickBooks environments into faster, auditable company-wide P&L reporting, targeting month-end close improvement from 10 days to 3 days.

Multi-facility scheduling & resource optimization

Optimize multi-facility league scheduling and resource allocation, reducing scheduling effort from 20 hours to 4 hours per week per facility while protecting revenue at risk.

Self-service practice booking

Enable 24/7 self-service practice booking in under 3 minutes, with a target of 35% self-service adoption and better capture of after-hours demand.

Architecture

From ad-hoc AI to governed, auditable AWS agents

The assessment validated an AWS-native architecture for governed agentic AI while aligning with LPS’s planned database modernization. Existing ad-hoc ChatGPT workflows were redesigned around secure, auditable controls, human-in-the-loop gates, PII handling, and long-term observability.

AWS services in the target architecture: Amazon Bedrock AgentCore (Runtime, Memory, Identity, Policy, Gateway, Observability, Evaluations), Amazon Quick Suite (Quick Index, Quick Flows, Quick Sight), Amazon Bedrock Guardrails, Amazon RDS, Amazon S3, Amazon QuickSight, Amazon SNS/SES, Amazon Cognito, IAM Identity Center, and AWS Secrets Manager.

The roadmap also supports the planned migration from SQL Server 2014 colocation to AWS RDS, helping create a centralized foundation for future AI-driven operations and ESOP/SOX-oriented auditability.

Results

Before and After

Metric Before After / Projected
Month-end close 10 days 3 days
Scheduling effort 20 hours per week per facility 4 hours per week per facility, an 80% reduction
Resume review time 15 minutes per application 2 minutes per application, an 87% reduction
Time to first interview 8 days 3 days
Self-service booking 0% self-service booking 35% adoption target with 24/7 booking in under 3 minutes
Schedule satisfaction 6.2 out of 10 8.5 out of 10 target
Scheduling revenue impact Up to 75,000 per year in revenue at risk $100,000 per year projected value from scheduling optimisation
AI governance Ad-hoc ChatGPT use with no formal guardrails Governed agents with guardrails, HITL gates, PII redaction and audit logs
Total annual impact Fragmented manual workflows across 19 facilities $100,000 projected annual savings across a 4-agent portfolio

Business impact

Projected impact across a four-agent portfolio

Phase 4 of the assessment produced a quantified business case with $100,000 in projected annual savings across four agents, against a $50K build investment, with a 1.5-month payback and 400% projected two-year ROI.

dollars in projected annual savings across 4 agents

dollars annual value from scheduling optimization

Month projected payback period

%

Reduction in manual scheduling labor

Partnering with AllCode allowed us to transform fragmented operational processes into a strategic asset. By modernizing our workflows with generative AI, we’re building the auditability and scalability required to meet our ESOP readiness goals with confidence and saving costs.

Gary Archer

CEO, Let’s Play Sports