# Consulting Service Catalog

Buyer-facing service menu for Edxperimental Labs consulting. Use this to route a prospect to the smallest sprint that can answer the actual AI decision.

| Service | Owner | Readiness | Buyer question | Next action |
| --- | --- | ---: | --- | --- |
| AI workflow benchmarking | Sanjay Prasad | 82 | Which model, provider, or agent route should handle this workflow? | Send one workflow and two candidate routes for a diagnostic scope. |
| Agent reliability review | Sanjay Prasad | 78 | Can this agent complete work safely across tools, browser state, and handoff boundaries? | Share a current agent demo, transcript, or run log for trace review. |
| Model and provider selection | Sanjay Prasad | 80 | What should run on frontier APIs, faster hosted models, open-weight inference, or human review? | Send workload volume, context length, output length, and candidate vendors. |
| AI security and risk sprint | Sanjay Prasad and Saujas | 74 | Where can prompt injection, excessive agency, data exposure, or weak escalation break the workflow? | Share the riskiest tool/action path and the data the agent should never reveal. |
| Sales-engineering diagnostic | Saujas | 86 | What is the smallest sprint that would answer the buyer's AI decision? | Send a short buyer problem statement and target decision date. |

## Service Details

### AI workflow benchmarking

- Owner: Sanjay Prasad
- Best for: Teams with one concrete process, examples, and a decision deadline.
- Starting inputs: Workflow examples; Expected outputs; Candidate systems; Failure cost
- Artifacts: Task packet; Run table; Trace ledger; Decision memo
- Readiness: 82/100
- Next action: Send one workflow and two candidate routes for a diagnostic scope.

### Agent reliability review

- Owner: Sanjay Prasad
- Best for: Teams piloting coding agents, browser agents, support agents, or internal automation.
- Starting inputs: Agent trace; Tool permissions; Success criteria; Human handoff rule
- Artifacts: Reliability scorecard; Failure taxonomy; Tool-risk map; Release gate
- Readiness: 78/100
- Next action: Share a current agent demo, transcript, or run log for trace review.

### Model and provider selection

- Owner: Sanjay Prasad
- Best for: Teams balancing quality, latency, cost, data boundary, and vendor risk.
- Starting inputs: Monthly volume; Latency target; Privacy boundary; Provider shortlist
- Artifacts: Route matrix; Cost curve; Fallback policy; Procurement memo
- Readiness: 80/100
- Next action: Send workload volume, context length, output length, and candidate vendors.

### AI security and risk sprint

- Owner: Sanjay Prasad and Saujas
- Best for: Teams moving from demo to production with tool access, customer data, or policy-sensitive outputs.
- Starting inputs: Threat model; Tool scope; Sensitive fields; Incident examples
- Artifacts: Security task pack; Risk register; Control deck; Launch blockers
- Readiness: 74/100
- Next action: Share the riskiest tool/action path and the data the agent should never reveal.

### Sales-engineering diagnostic

- Owner: Saujas
- Best for: Founders or operators who need a scoped benchmark before a larger AI build.
- Starting inputs: Business goal; Stakeholder map; Current workflow; Budget signal
- Artifacts: Discovery memo; Sprint scope; Access checklist; Proposal outline
- Readiness: 86/100
- Next action: Send a short buyer problem statement and target decision date.