01 · Install
Configure an evaluation-only, fail-closed interceptor path alongside the existing WSO2 baseline and Prompt Compressor path.
Independent WSO2 compatibilityFixed-scope evaluation · $5,000
Reduce AI context inside WSO2. Keep the evidence path.
Evaluate Maha Context Compiler as a fail-closed request interceptor in an existing WSO2 AI Gateway deployment—without replacing WSO2 as the enterprise gateway and control plane.
The pilot measures provider input tokens, required-fact retention, citation traceability, latency, failure behavior, and model cost against a customer-shaped workload. It returns a private evidence package and a recommendation to proceed, revise, or stop.
02 · Compare
Run one agreed workload through all three paths with a frozen configuration, explicit cost ceiling, and no automatic retries.
03 · Decide
Deliver aggregate results, sanitized traces, failure-path evidence, limitations, and a production-readiness recommendation.
Preliminary observed result
A reason to evaluate—not a universal performance claim.
In a 60-call comparison over 20 frozen, synthetic 20K–100K-token workloads, the Maha path reduced provider input tokens by 98.84% and observed model cost by 98.20% relative to baseline. All calls completed without retries or failures.
| Path | Provider input tokens | Observed model cost | Required facts retained |
|---|---|---|---|
| WSO2 baseline | 1,621,553 | $1.632963 | 60 / 60 |
| WSO2 Prompt Compressor | 1,489,323 | $1.505248 | 0 / 60* |
| WSO2 + Maha | 18,849 | $0.029379 | 60 / 60 |
*The Prompt Compressor result is specific to WSO2 AI Gateway 1.1.0, Prompt Compressor 0.9.0, and a 0.55 retained ratio. It should not be generalized before WSO2 or a customer confirms that configuration reflects the intended production setup. Provider pricing assumptions and the full frozen configuration are recorded in the reproduction manifest.
Commercial pilot
One workload. One fixed fee. A decision at the end.
Standard bounded evaluation
$5,000
- • One sanitized, customer-shaped document or RAG workflow
- • Up to 1,000 test requests within an agreed provider-spend ceiling
- • Baseline, WSO2 Prompt Compressor, and Maha comparison
- • Retention, citations, latency, cost, and fail-closed analysis
- • Private findings review and production recommendation
If the result is useful
Continue with a hosted request-based plan or an annual private-deployment license. Commercial terms are proposed only after the pilot establishes an actual workload advantage.
Founding design-partner evaluations may be scoped at $2,500 when the customer can provide structured technical feedback and permits an anonymized integration note.
Reproduce before trusting
The evaluation method is inspectable.
The public package pins the gateway and policy versions, retained ratio, model, corpus digest, failure behavior, and zero-retry rule. Its default one-command run is a dry run and makes no provider calls.
Evaluation policy bundle ↗Frozen reproduction manifest ↗Sanitized three-path trace ↗Technical integration notes ↗
npm run reproduce:wso2-evaluationDeclared boundaries
Independent compatibility work, not a WSO2 endorsement.
- • Maha Strategies is not claiming WSO2 partnership, certification, approval, or customer validation.
- • The published benchmark corpus is synthetic; the pilot exists to test whether the result survives a realistic customer workload.
- • The public policy bundle is evaluation-only. Production requires a reviewed gateway-side secret reference, service identity, mTLS, or equivalent interceptor authentication.
- • No fixed compression, savings, retention, or latency result is promised before measurement.
Start with evidence
Bring one expensive or audit-sensitive context workflow.
Send the approximate input size, request volume, current WSO2 AI Gateway version, and the facts or citations that must survive. Maha will reply with a bounded scope—or say plainly if the workflow is not a fit.