Inject

Any data item

Connectors or approved one-off imports load email addresses, cookies, events, CRM rows, products, support, billing, documents, and outcomes.

Store

Attribution paths

Scout stores each item's relationship set, vector, and ordered attribution path in customer-owned PostgreSQL/pgvector. Fortress and Elite move high-load memory into Rust/LanceDB, not KynticAI product operations.

Compare

Rust + LanceDB

Enterprise contains the proprietary Rust relationship, weighting, traversal, and LanceDB path for fast similarity analysis across millions of relationship sets.

Output

Top-example JSON

The engine creates JSON with the best examples, weights, confidence, caveats, and ranked task options for the question being asked.

Explain

LLM task brief

Fortress sends JSON to your chosen LLM. Elite can use KynticAI's on-prem open-source LLM path with no third-party token charges.

Back to case studies
NHS / healthcare operationsICB operations lead

Non-clinical pathway pressure and protected capacity action

Referral backlog, clinic capacity, transport, and admin blockers become an operations recommendation.

A pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. The pathway team needs a non-clinical action that protects operational capacity without becoming clinical decision support.

Runtime Query

The scenario is run as a query against comparable relationship sets.

A non-clinical pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. What operations action should the pathway team take next?

Run 2026-06-17-base-run / nhs-healthcare-operations/query-results/urgent-risk-next-action.json

pathway_capacity_protected

Primary outcome

Highest-ranked outcome from the runtime query result.

0.7118

Best score

Runtime score for the primary action option.

3

Options

Primary plus alternative action options returned by the runtime.

Data Examples

  • Referral backlog growth signal
  • Unfilled clinic capacity sessions
  • Transport delay coordination note
  • Admin validation queue with missing contact confirmation
  • Comparable protected-capacity, admin-blocker, and patient-comms outcomes

Event Path

  1. 01referral_backlog_alert
  2. 02clinic_capacity_gap
  3. 03patient_transport_delay
  4. 04admin_validation_blocker
  5. 05capacity_huddle_opened
  6. 06pathway_capacity_protected

Scout Output

  • Runtime generated Scout-shaped data items for the domain.
  • Runtime generated Scout relationships and ordered attribution paths.
  • Domain manifest cross-checks the journey, event, and relationship-set counts.

Fortress Output

  • Runtime generated comparable relationship sets.
  • Runtime returned ranked action options with scores, confidence labels, outcomes, and supporting relationship-set IDs.
  • Runtime preserved synthetic-only and customer-data false caveats.

Ranked Outcomes

Fortress returns a primary recommendation and the credible alternatives.

Primary Recommendation

Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.

Outcomepathway_capacity_protected

ConfidenceHigh

Score0.7118

This shows non-clinical operational intelligence: capacity huddle, validated backlog list, and protected slot pool.

Other Ranked Options

Option 2 - Medium - score 0.3698

Prioritise admin validation and contact-detail correction before requesting more pathway capacity.

Expected outcome: admin_blocker_resolved

Option 3 - Medium - score 0.3106

Send the approved non-clinical appointment communication and rebooking route before escalating capacity.

Expected outcome: patient_comms_sent

Elite LLM Handoff

Explain the reduce_pathway_pressure recommendation using the governed Fortress output and caveats.

governed_handoff:
  - Goal: reduce_pathway_pressure
  - Primary path: pathway_capacity_protected, score 0.7118, high confidence
  - Supporting signals: admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay
  - Runtime relationship sets returned: 8
  - Domain journeys: 120; events: 600; relationship sets: 120

Outcome Prompt JSON

{
  "schema": "kynticai.elite.outcome-prompts.v1",
  "caseStudyId": "nhs-healthcare-operations-pathway-pressure",
  "domain": "NHS / healthcare operations",
  "stakeholder": "ICB operations lead",
  "runtimeQuery": "A non-clinical pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. What operations action should the pathway team take next?",
  "promptContract": "You are Elite, the KynticAI recommendation-writing layer.\n\nYou receive governed Fortress output, not raw customer data. Use only the supplied relationship-set facts, scores, confidence labels, alternatives, caveats, and outcome goal.\n\nWrite a decisive operational recommendation for the accountable team. Your output must include:\n1. Recommended action.\n2. Why this action is supported by the comparable relationship sets.\n3. What to do now.\n4. When to choose one of the alternative actions.\n5. What outcome to capture so the system improves.\n\nRules:\n- Do not invent facts outside the governed handoff.\n- Do not promise revenue, retention, clinical, legal, financial, safety, or compliance outcomes.\n- Do not apologise for using synthetic demo data.\n- Be commercially useful, specific, and direct.\n- Keep human review and claim boundaries visible where the domain requires them.",
  "governedFacts": [
    "Goal: reduce_pathway_pressure",
    "Primary path: pathway_capacity_protected, score 0.7118, high confidence",
    "Supporting signals: admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay",
    "Runtime relationship sets returned: 8",
    "Domain journeys: 120; events: 600; relationship sets: 120"
  ],
  "outcomePrompts": [
    {
      "rank": 1,
      "outcome": "pathway_capacity_protected",
      "action": "Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.",
      "confidence": "High",
      "score": "0.7118",
      "prompt": "You are Elite, the KynticAI recommendation-writing layer.\n\nUse only the governed Fortress handoff below. Do not invent facts.\n\nCase study: Non-clinical pathway pressure and protected capacity action\nDomain: NHS / healthcare operations\nStakeholder: ICB operations lead\nRuntime query: A non-clinical pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. What operations action should the pathway team take next?\n\nGoverned facts:\n- Goal: reduce_pathway_pressure\n- Primary path: pathway_capacity_protected, score 0.7118, high confidence\n- Supporting signals: admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay\n- Runtime relationship sets returned: 8\n- Domain journeys: 120; events: 600; relationship sets: 120\n\nCandidate outcome to explain:\n- Rank: 1\n- Outcome: pathway_capacity_protected\n- Action: Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.\n- Confidence: High\n- Score: 0.7118\n- Evidence summary: Comparable relationship sets ended in pathway_capacity_protected and share signals such as admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay.\n- Supporting relationship sets: set-f7952eb1243b, set-18442a874900, set-21fc983fc52e, set-0677db0d3073, set-af13652b74f6\n\nWrite the operational recommendation for this outcome.\nInclude why the comparable relationship sets support it, what the team should do now, when this outcome should be chosen over the other ranked options, and what reviewed outcome must be captured next.",
      "expectedTextOutput": "Recommended action: Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.\n\nWhy this is supported: Fortress ranked pathway_capacity_protected first with score 0.7118 and high confidence. Comparable relationship sets ended in pathway_capacity_protected and share signals such as admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay. The supporting relationship sets are set-f7952eb1243b, set-18442a874900, set-21fc983fc52e, set-0677db0d3073, set-af13652b74f6.\n\nWhat to do now: brief the icb operations lead with the runtime query, the shared signals (admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay), and the exact action. Keep the action specific to this scenario rather than turning it into a generic campaign or workflow rule.\n\nWhen to choose an alternative: switch if the live evidence is closer to admin_blocker_resolved (Prioritise admin validation and contact-detail correction before requesting more pathway capacity.); patient_comms_sent (Send the approved non-clinical appointment communication and rebooking route before escalating capacity.).\n\nOutcome to capture: record the action taken, the reviewed outcome, and any contradictory signal so the next Scout/Fortress comparison has stronger evidence."
    },
    {
      "rank": 2,
      "outcome": "admin_blocker_resolved",
      "action": "Prioritise admin validation and contact-detail correction before requesting more pathway capacity.",
      "confidence": "Medium",
      "score": "0.3698",
      "prompt": "You are Elite, the KynticAI recommendation-writing layer.\n\nUse only the governed Fortress handoff below. Do not invent facts.\n\nCase study: Non-clinical pathway pressure and protected capacity action\nDomain: NHS / healthcare operations\nStakeholder: ICB operations lead\nRuntime query: A non-clinical pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. What operations action should the pathway team take next?\n\nGoverned facts:\n- Goal: reduce_pathway_pressure\n- Primary path: pathway_capacity_protected, score 0.7118, high confidence\n- Supporting signals: admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay\n- Runtime relationship sets returned: 8\n- Domain journeys: 120; events: 600; relationship sets: 120\n\nCandidate outcome to explain:\n- Rank: 2\n- Outcome: admin_blocker_resolved\n- Action: Prioritise admin validation and contact-detail correction before requesting more pathway capacity.\n- Confidence: Medium\n- Score: 0.3698\n- Evidence summary: Comparable relationship sets ended in admin_blocker_resolved and share signals such as blocker_resolved, booking_admin, contact_detail_gap, validation_queue.\n- Supporting relationship sets: set-ebeecd48a1dc\n\nWrite the operational recommendation for this outcome.\nInclude why the comparable relationship sets support it, what the team should do now, when this outcome should be chosen over the other ranked options, and what reviewed outcome must be captured next.",
      "expectedTextOutput": "Alternative action: Prioritise admin validation and contact-detail correction before requesting more pathway capacity.\n\nWhy this route exists: Fortress returned admin_blocker_resolved at rank 2 with score 0.3698 and medium confidence. Comparable relationship sets ended in admin_blocker_resolved and share signals such as blocker_resolved, booking_admin, contact_detail_gap, validation_queue. The supporting relationship sets are set-ebeecd48a1dc.\n\nWhen to use it: choose this instead of the primary route when the live evidence matches blocker_resolved, booking_admin, contact_detail_gap, validation_queue more closely than the current top-ranked outcome.\n\nWhat to do now: make the action explicit for the icb operations lead, preserve the caveat for this domain, and avoid inventing extra facts beyond the governed handoff.\n\nOutcome to capture: record whether admin_blocker_resolved happened and feed that reviewed result back into the relationship-set evidence."
    },
    {
      "rank": 3,
      "outcome": "patient_comms_sent",
      "action": "Send the approved non-clinical appointment communication and rebooking route before escalating capacity.",
      "confidence": "Medium",
      "score": "0.3106",
      "prompt": "You are Elite, the KynticAI recommendation-writing layer.\n\nUse only the governed Fortress handoff below. Do not invent facts.\n\nCase study: Non-clinical pathway pressure and protected capacity action\nDomain: NHS / healthcare operations\nStakeholder: ICB operations lead\nRuntime query: A non-clinical pathway has referral backlog pressure, clinic capacity gaps, patient transport delays, and admin validation blockers. What operations action should the pathway team take next?\n\nGoverned facts:\n- Goal: reduce_pathway_pressure\n- Primary path: pathway_capacity_protected, score 0.7118, high confidence\n- Supporting signals: admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay\n- Runtime relationship sets returned: 8\n- Domain journeys: 120; events: 600; relationship sets: 120\n\nCandidate outcome to explain:\n- Rank: 3\n- Outcome: patient_comms_sent\n- Action: Send the approved non-clinical appointment communication and rebooking route before escalating capacity.\n- Confidence: Medium\n- Score: 0.3106\n- Evidence summary: Comparable relationship sets ended in patient_comms_sent and share signals such as appointment_reminder_gap, approved_comms, attendance_protected, rebooking_route.\n- Supporting relationship sets: set-ea253d51fe99\n\nWrite the operational recommendation for this outcome.\nInclude why the comparable relationship sets support it, what the team should do now, when this outcome should be chosen over the other ranked options, and what reviewed outcome must be captured next.",
      "expectedTextOutput": "Alternative action: Send the approved non-clinical appointment communication and rebooking route before escalating capacity.\n\nWhy this route exists: Fortress returned patient_comms_sent at rank 3 with score 0.3106 and medium confidence. Comparable relationship sets ended in patient_comms_sent and share signals such as appointment_reminder_gap, approved_comms, attendance_protected, rebooking_route. The supporting relationship sets are set-ea253d51fe99.\n\nWhen to use it: choose this instead of the primary route when the live evidence matches appointment_reminder_gap, approved_comms, attendance_protected, rebooking_route more closely than the current top-ranked outcome.\n\nWhat to do now: make the action explicit for the icb operations lead, preserve the caveat for this domain, and avoid inventing extra facts beyond the governed handoff.\n\nOutcome to capture: record whether patient_comms_sent happened and feed that reviewed result back into the relationship-set evidence."
    }
  ]
}

Elite Textual Recommendation

Recommended action: Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.

Why this is supported: Fortress ranked pathway_capacity_protected first with score 0.7118 and high confidence. Comparable relationship sets ended in pathway_capacity_protected and share signals such as admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay. The supporting relationship sets are set-f7952eb1243b, set-18442a874900, set-21fc983fc52e, set-0677db0d3073, set-af13652b74f6.

What to do now: brief the icb operations lead with the runtime query, the shared signals (admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay), and the exact action. Keep the action specific to this scenario rather than turning it into a generic campaign or workflow rule.

When to choose an alternative: switch if the live evidence is closer to admin_blocker_resolved (Prioritise admin validation and contact-detail correction before requesting more pathway capacity.); patient_comms_sent (Send the approved non-clinical appointment communication and rebooking route before escalating capacity.).

Outcome to capture: record the action taken, the reviewed outcome, and any contradictory signal so the next Scout/Fortress comparison has stronger evidence.

Generated Elite Action Brief

  1. 01Recommended action: Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.
  2. 02Evidence: Comparable relationship sets ended in pathway_capacity_protected and share signals such as admin_validation_blocker, capacity_huddle, clinic_capacity_gap, referral_backlog, transport_delay.
  3. 03Primary action: Open a non-clinical capacity huddle, validate the backlog list, and move suitable appointments into the protected slot pool for operational review.
  4. 04Primary expected outcome: pathway_capacity_protected
  5. 05Alternatives: admin_blocker_resolved: Prioritise admin validation and contact-detail correction before requesting more pathway capacity. | patient_comms_sent: Send the approved non-clinical appointment communication and rebooking route before escalating capacity.
  6. 06Capture the reviewed outcome and feed it back into the relationship-set evidence.

Reviewed outcome loop

  • Capture which ranked action was taken.
  • Record whether the expected outcome happened.
  • Feed the reviewed outcome back into the next relationship-set comparison.

Boundary

Synthetic non-clinical operations demo only. It is not clinical advice, patient triage, diagnosis, treatment, or NHS deployment evidence.

Evidence Results

Case studies show the same relationship engine in different workflows.

These examples help buyers recognise the pattern in their own sales, support, operational, or governance data.

KynticAI Result
Case Studies

Runtime pack - multiple domains, one relationship-analysis pattern

Recognise your own workflow

The case studies prove the same product chain across ecommerce, logistics, legal, manufacturing, education, healthcare operations, and professional services.
KynticAI Result
Elite Output

Runtime pack - governed JSON to text recommendation

See the model explain evidence

Each case shows how Elite turns governed Fortress JSON into a clear operational recommendation.
KynticAI Result
Outcome Capture

Runtime pack - feedback into relationship memory

Turn examples into a learning system

Each case ends by asking what outcome should be captured so the recommendation loop improves.