Persona Chat Simulation API

How it works

List catalog

Authorization: Bearer sk_live_…
X-Enterprise-Name: suki

Use clinicians[].id, patientPopulations[].id, and scenario keys when you start a job. If summary / traits is null, that population’s description was unavailable. Browse every population and its traits below; the JSON example under it is a real C Section sample with full summary and traits.

{
  "enterpriseName": "suki",
  "clinicians": [
    {
      "id": "9512b843-9b5c-4295-808c-08daa17d4705",
      "name": "Medsurge Nurse"
    }
  ],
  "patientPopulations": [
    {
      "id": "35e7f35d-a029-4639-91a6-fc4d5fdc668c",
      "name": "C Section",
      "summary": "This group represents people navigating urgent, high-stakes life transitions centered on pregnancy, childbirth, and early parenting—particularly those facing medical decisions around cesarean sections, postpartum recovery, infant feeding, and managing the physical and emotional aftermath of traumatic or complicated births. They arrive at online communities seeking concrete, experience-based guidance during moments of acute uncertainty: choosing between elective C-section and vaginal birth, troubleshooting breastfeeding or pumping challenges, managing surgical recovery while caring for newborns, and making rapid decisions about infant health and safety. Success for them means finding actionable solutions that reduce physical pain, protect their babies, preserve their mental health, and help them feel prepared and in control during an overwhelming period when medical advice alone often feels insufficient or dismissive.\n\nThese users operate under severe constraints: sleep deprivation, physical pain and mobility limits, time scarcity due to newborn care demands, financial pressures around childcare and medical costs, limited local support networks, and the need to make irreversible medical decisions quickly while emotionally and physically depleted. Many are first-time parents or facing complications that make past experience less relevant, and they frequently mention partners, family members, or healthcare providers who are unsupportive, unavailable, or inconsistent, forcing them to become their own advocates and researchers. Geographic and systemic barriers—rural locations, language differences, insurance limitations, hospital policies—further constrain their options and amplify their need for peer-sourced workarounds and validation.\n\nTheir decision-making is intensely pragmatic and experience-driven: they prioritize firsthand accounts over abstract medical guidelines, ask for specific timelines and product names, cross-post the same question across multiple communities to triangulate answers, and explicitly request stories from people \"like them\" to assess realistic tradeoffs. They conduct moderate research—reading threads, watching for patterns in replies, consulting professionals when accessible—but ultimately rely heavily on lived experience because clinical advice often feels generic, dismissive of their concerns, or fails to address the practical realities of recovery and caregiving. They favor incremental, reversible solutions they can try at home, but will escalate to specialists or emergency care when peer consensus or their own instincts signal real risk.\n\nEmotionally, this group oscillates between anxiety and determination: they express fear about pain, trauma recurrence, and making the \"wrong\" choice, yet they also demonstrate remarkable agency—switching providers, demanding second opinions, setting firm boundaries with family, and persistently seeking answers until they feel heard. Cognitive load is high and visible in their posts: they share detailed medical histories, track symptoms and timelines obsessively, ask for reassurance repeatedly, and often admit to feeling overwhelmed, guilty, or isolated. Many describe past trauma—emergency surgeries, dismissive clinicians, severe complications—that shapes their current risk calculus and makes them hyper-vigilant about safety and bodily autonomy.\n\nTrust is earned through specificity and solidarity: they value commenters who share concrete outcomes (recovery timelines, medication regimens, product brands), disclose similar complications, and validate their feelings without judgment. They distrust vague reassurances, one-size-fits-all recommendations, and advice that minimizes their pain or fear, and they are quick to call out perceived conflicts of interest, performative content, or responses that feel like marketing. Risk concerns dominate their discussions—infection, hemorrhage, supply loss, infant safety, re-traumatization, and making a choice they'll regret—and they repeatedly seek confirmation that their symptoms, fears, or plans fall within a \"normal\" range or justify urgent action.\n\nOn forums, they ask highly specific, situational questions: \"Has anyone had a C-section at [hospital]?\", \"What did your scar look like at 3 weeks?\", \"How soon can I lift my toddler?\", \"Which pump flanges worked for you?\" They offer detailed context to get relevant answers, follow up with gratitude and outcomes, and frequently pay forward the help they received by answering newer members' questions with the same specificity. Frustrations voiced include dismissive or inconsistent medical advice, lack of local resources, partners or family who don't understand the physical toll, and the gap between idealized parenting narratives and their messy, painful realities.\n\nIn a marketplace context, these users would exhibit distinct behavioral patterns: they need solutions that are immediately actionable and require minimal cognitive load to evaluate; they will read reviews obsessively and prioritize peer testimonials over brand claims; they are willing to pay for convenience, comfort, or peace of mind but are also budget-conscious and will seek secondhand or multi-use options; they value fast shipping, easy returns, and clear instructions because they have no margin for trial-and-error; and they will abandon purchases or platforms that feel predatory, opaque, or dismissive of their urgent, specific needs. They are highly motivated buyers in a narrow window—desperate for relief and solutions—but also protective, skeptical, and quick to share warnings about products or services that failed them, making authenticity and transparency non-negotiable for earning their trust and repeat engagement.",
      "traits": [
        {
          "title": "Situational Triggers & Context",
          "keywordTags": [
            "Postpartum recovery urgency",
            "High-stakes medical decisions",
            "Time-sensitive infant care",
            "Boundary-setting with family",
            "Navigating healthcare systems"
          ],
          "descriptions": [
            "Users are triggered into marketplace activity by immediate postpartum physical challenges—C-section wound care, pain management, mobility limitations, breastfeeding complications—that require quick product solutions (belly binders, peri bottles, nipple shields, pumps) to restore function and comfort.",
            "The decision to pursue elective C-section, VBAC, or repeat surgery creates intense information-seeking about recovery timelines, scar outcomes, anesthesia options, and hospital policies, often driven by prior birth trauma or fear of emergency interventions.",
            "Urgent infant needs—feeding difficulties, sleep deprivation, reflux, jaundice, supply concerns—push users to seek gear, formula, bottles, or services (lactation consultants, night nannies) that promise immediate relief or measurable improvement.",
            "Conflict with in-laws, partners, or medical providers over visitation, feeding choices, or care decisions prompts users to seek validation, boundary scripts, and sometimes physical solutions (nursery cameras, separate sleeping spaces) to enforce autonomy.",
            "Navigating fragmented or dismissive healthcare (insurance coding disputes, delayed specialist referrals, conflicting provider advice) drives users to crowdsource workarounds, second opinions, and concrete next steps to regain control and clarity."
          ],
          "confidenceScore": 82,
          "confidenceReasoning": "This trait is strongly supported by explicit, recurring situational details across the majority of profiles—users consistently describe specific medical events, recovery phases, infant ages, and interpersonal conflicts as the immediate context for their engagement. The high frequency and specificity of these triggers (weeks postpartum, gestational age, procedure type) provide clear behavioral anchors. Confidence is slightly reduced because some profiles focus more on the advice-giving side and reveal less about their own current situational pressures, and there is limited visibility into how triggers evolve as the postpartum period progresses beyond the first few months."
        },
        {
          "title": "Decision Drivers & Tradeoffs",
          "keywordTags": [
            "Safety and infant wellbeing first",
            "Maternal physical and mental health preservation",
            "Predictability over idealized outcomes",
            "Cost-conscious pragmatism",
            "Partner and family support availability"
          ],
          "descriptions": [
            "Users consistently prioritize infant safety, health, and comfort as the non-negotiable anchor—choosing products, procedures, or services that minimize risk (safe sleep, feeding adequacy, infection prevention) even when those choices are more expensive, effortful, or socially stigmatized.",
            "Maternal recovery, pain management, mental health, and functional capacity are weighed heavily in tradeoffs; users will choose formula over breastfeeding, hire help, or select elective surgery explicitly to preserve their own wellbeing and ability to care for their child, rejecting guilt or external pressure.",
            "Predictability, control, and avoiding worst-case scenarios (emergency C-section, severe tearing, hemorrhage, supply failure) often outweigh pursuit of 'natural' or idealized birth and parenting experiences; users favor planned interventions and known timelines to reduce anxiety and trauma risk.",
            "Budget constraints and cost-effectiveness are frequently cited—users seek secondhand gear, compare formula prices, weigh insurance coverage, and balance spending on essentials (reliable car seats, effective pumps) against optional or luxury items, showing resourcefulness and frugality.",
            "Availability of partner support, family help, or paid assistance (doulas, night nannies, housekeepers) is a critical tradeoff factor; users adjust plans based on whether they will be alone during recovery, have reliable backup, or need to hire services to fill gaps in their support network."
          ],
          "confidenceScore": 85,
          "confidenceReasoning": "This trait is robustly evidenced by explicit tradeoff discussions across nearly all profiles—users routinely articulate competing priorities (safety vs. cost, mental health vs. breastfeeding ideals, control vs. convenience) and explain their reasoning in concrete terms. The consistency of these drivers (infant safety, maternal wellbeing, predictability) and the frequency with which users name specific tradeoffs (e.g., 'I chose formula to preserve my sanity,' 'elective C-section for control') provide strong behavioral signals. Confidence is not higher because the relative weighting of these drivers can vary by individual and context, and some profiles show more emphasis on one driver (e.g., cost) than others, suggesting heterogeneity that could affect prediction accuracy in edge cases."
        },
        {
          "title": "Risk, Trust & Verification Behavior",
          "keywordTags": [
            "Skepticism of dismissive providers",
            "Peer experience as primary evidence",
            "Escalation when concerns are minimized",
            "Distrust of marketing and influencers",
            "Preference for measurable outcomes"
          ],
          "descriptions": [
            "Users express distrust toward medical providers who dismiss symptoms, breach consent, or fail to communicate risks clearly, and they actively seek second opinions, switch providers, or escalate to specialists (MFM, pelvic PT, lactation consultants) when initial care feels inadequate or coercive.",
            "Lived experience from peers—detailed birth stories, recovery timelines, product performance, feeding outcomes—is treated as more credible and actionable than abstract clinical reassurances or provider platitudes; users explicitly request 'what actually happened to you' rather than general advice.",
            "When symptoms worsen, bleeding is abnormal, or pain is severe, users are quick to escalate to triage, ER, or urgent consultations, showing a risk-averse, safety-first stance that prioritizes professional confirmation over waiting or self-managing uncertainty.",
            "Users are skeptical of baby-product marketing, influencer endorsements, registry checklists, and 'must-have' lists, frequently calling out overconsumption, performative parenting content, and predatory postpartum services, preferring evidence-based or peer-tested recommendations.",
            "Concrete, measurable indicators—HCG doubling times, baby weight percentiles, milk output volumes, scar measurements, pain scales—are used to verify progress and assess risk, reflecting a preference for objective data over subjective reassurance when stakes are high."
          ],
          "confidenceScore": 78,
          "confidenceReasoning": "This trait is well-supported by recurring patterns of skepticism, peer-reliance, and escalation behavior across many profiles, with users explicitly describing distrust of providers, preference for community testimony, and use of measurable markers. However, the evidence is somewhat more implicit in some profiles (inferred from advice-giving rather than direct statements of personal distrust), and there is variability in how aggressively users escalate or switch providers—some are more deferential or constrained by access. The trait is also more visible in medical and product contexts than in service or interpersonal trust decisions. Additional signals about trust repair, when users do trust providers, and how they verify non-medical claims would strengthen confidence."
        },
        {
          "title": "Information-Seeking & Evaluation Style",
          "keywordTags": [
            "Crowdsourcing lived experiences",
            "Iterative trial-and-error",
            "Targeted, specific questions",
            "Cross-posting for triangulation",
            "Preference for actionable steps over theory"
          ],
          "descriptions": [
            "Users systematically crowdsource firsthand accounts by posting detailed, context-rich questions across multiple relevant subreddits (local, medical, product-specific) to gather diverse peer experiences and triangulate advice before deciding, showing a deliberate, community-reliant research strategy.",
            "Decision-making is iterative and experimental—users try products, routines, or interventions in small steps, monitor outcomes closely (baby's response, their own symptoms, measurable changes), and adjust or abandon approaches based on real-world results rather than committing to a single plan upfront.",
            "Questions are highly specific and situational, often including precise details (weeks postpartum, baby's age, exact symptoms, product models, local constraints) to elicit relevant, comparable experiences rather than generic advice, indicating a need for tailored, applicable information.",
            "Users frequently cross-post the same question to multiple communities, follow up on threads to request updates or outcomes, and synthesize responses to identify consensus or outliers, demonstrating a methodical approach to reducing uncertainty through volume and comparison.",
            "Actionable, step-by-step guidance—product links, brand names, procedural instructions, timelines, concrete next steps—is strongly preferred over abstract explanations, theoretical frameworks, or reassurances without practical application, reflecting high cognitive load and urgency for implementable solutions."
          ],
          "confidenceScore": 87,
          "confidenceReasoning": "This trait is very strongly supported by observable behavior patterns across the dataset—users explicitly cross-post, request specific details, share iterative experiments, and thank responders for concrete tips. The consistency of these information-seeking behaviors (crowdsourcing, specificity, actionability) and the frequency with which users describe their research process provide robust evidence. Confidence is high but not maximal because there is some variation in how systematically users research (some rely more on quick community consensus, others do deeper dives), and the profiles capture advice-giving more than the full arc of information-gathering and synthesis. Additional visibility into how users weigh conflicting advice and decide when they have 'enough' information would further strengthen this trait."
        },
        {
          "title": "Emotional State & Cognitive Load",
          "keywordTags": [
            "Sleep-deprived and overwhelmed",
            "Anxiety about worst-case outcomes",
            "Guilt and self-judgment",
            "Desire for validation and normalization",
            "Protective urgency for infant"
          ],
          "descriptions": [
            "Users frequently describe severe sleep deprivation, exhaustion, brain fog, and feeling overwhelmed by the volume of decisions and tasks, which drives preference for simple, low-friction solutions and short, digestible advice rather than complex research or multi-step processes.",
            "Anxiety about catastrophic outcomes—hemorrhage, infection, supply failure, developmental delays, SIDS, birth trauma recurrence—is pervasive and shapes risk-averse, precautionary behavior, with users seeking reassurance but also concrete safety measures to mitigate feared scenarios.",
            "Guilt, self-blame, and fear of judgment (for feeding choices, birth outcomes, asking for help, not enjoying motherhood) are openly expressed, and users actively seek validation that their choices are acceptable, that struggles are normal, and that they are 'good enough' parents despite imperfections.",
            "Users crave normalization of difficult experiences—painful recovery, breastfeeding failure, postpartum mood struggles, partner conflict—and respond positively to peers who share similar stories, using community solidarity to reduce isolation and reframe challenges as common rather than personal failings.",
            "A fierce, protective urgency around infant safety and wellbeing drives rapid decision-making and low tolerance for perceived threats (unsafe products, dismissive providers, boundary-violating relatives), with users willing to act decisively—sometimes confrontationally—to shield their child from harm."
          ],
          "confidenceScore": 80,
          "confidenceReasoning": "This trait is strongly supported by explicit emotional disclosures across many profiles—users openly describe anxiety, exhaustion, guilt, and protective instincts, and these states clearly influence their behavior (preference for simple solutions, need for validation, rapid escalation). The consistency of these emotional patterns and their direct link to decision-making provide solid evidence. Confidence is slightly reduced because emotional intensity and cognitive load likely vary over time (acute postpartum vs. later months) and by individual resilience, and some profiles focus more on practical problem-solving with less emotional disclosure. Additionally, the profiles capture a snapshot rather than longitudinal emotional trajectories, so it's unclear how these states evolve or resolve, which limits predictive confidence for longer-term engagement."
        },
        {
          "title": "Platform Control & Guidance Preference",
          "keywordTags": [
            "Preference for peer testimony over algorithms",
            "Desire for curated, vetted recommendations",
            "Frustration with information overload",
            "Appreciation for step-by-step instructions",
            "Skepticism of sponsored or promotional content"
          ],
          "descriptions": [
            "Users actively seek out peer-driven communities (subreddits, forums, local groups) rather than relying on search engines, influencer feeds, or algorithm-curated content, valuing the authenticity, specificity, and accountability of real people sharing real experiences over polished or sponsored recommendations.",
            "There is strong demand for curated, vetted lists—trusted product recommendations, local provider referrals, evidence-based resources—that reduce decision fatigue and filter out marketing noise, with users frequently asking 'what actually worked for you' to bypass overconsumption and hype.",
            "Information overload and conflicting advice (from providers, family, online sources) cause frustration and paralysis; users express relief when someone offers a clear, consensus-backed answer or a simple decision framework that cuts through the noise and reduces cognitive burden.",
            "Step-by-step instructions, checklists, timelines, and procedural guides are highly valued—users share and request detailed how-tos (hospital bag lists, pump settings, scar-care routines, feeding schedules) that provide structure and reduce the need to synthesize fragmented information under stress.",
            "Users are alert to and critical of promotional content, affiliate links, influencer shilling, and registry-driven marketing, often calling out posts that feel like ads or that push unnecessary products, and they prefer transparent, non-commercial peer recommendations or explicitly disclosed sponsorships."
          ],
          "confidenceScore": 74,
          "confidenceReasoning": "This trait is moderately well-supported by observable preferences for peer communities, requests for specific recommendations, and skepticism toward marketing, but the evidence is more implicit than explicit—users demonstrate these preferences through their behavior (where they post, what they ask for) rather than directly articulating their platform or guidance preferences. There is clear frustration with information overload and appreciation for actionable steps, but less visibility into how users would respond to different platform features (e.g., expert curation vs. peer voting, personalized recommendations vs. open forums). The trait is also more evident in product and medical contexts than in service or community-building contexts. Additional direct feedback on platform design, moderation, and content curation would strengthen confidence."
        },
        {
          "title": "Commitment, Switching & Post-Decision Behavior",
          "keywordTags": [
            "Willingness to switch when dissatisfied",
            "Iterative refinement of choices",
            "Follow-up and outcome-sharing",
            "Loyalty to what works",
            "Advocacy and reciprocal helping"
          ],
          "descriptions": [
            "Users demonstrate low switching costs and high willingness to change providers, products, or plans when initial choices prove inadequate—switching OBs, trying different formulas or pumps, abandoning sleep methods—showing pragmatic flexibility and prioritization of outcomes over sunk costs or loyalty.",
            "Post-decision behavior is iterative rather than final; users continuously refine choices based on observed results (adjusting pump settings, tweaking routines, adding supplements), treating decisions as experiments to be optimized rather than one-time commitments, and they return to communities for troubleshooting.",
            "Users frequently follow up on threads to report outcomes, thank responders, and share what worked or didn't, creating a reciprocal information loop that helps future users; this follow-up behavior signals both gratitude and a desire to 'pay forward' the help they received.",
            "When a product, provider, or strategy works well, users become loyal advocates—repeatedly recommending specific brands, practitioners, or approaches across multiple threads and defending those choices when questioned, showing strong positive reinforcement and word-of-mouth behavior.",
            "Users actively contribute to community knowledge by answering others' questions, sharing detailed experiences, and offering emotional support, demonstrating a pattern of reciprocal helping and community investment that extends beyond their own immediate needs and reflects ongoing engagement with the peer network."
          ],
          "confidenceScore": 76,
          "confidenceReasoning": "This trait is well-supported by observable patterns of switching, iteration, follow-up, and advocacy across many profiles, with users explicitly describing changes in plans, reporting back on outcomes, and recommending what worked. The reciprocal helping behavior and loyalty to effective solutions are also clearly evident. However, the evidence is somewhat more visible in the advice-giving profiles than in the advice-seeking ones, and there is less clarity about the conditions under which users disengage entirely (stop posting, leave communities) versus continue iterating. The trait also assumes ongoing platform engagement, but some users may resolve their issues and leave without follow-up, which is not captured in the data. Additional signals about long-term engagement patterns, churn, and the triggers for disengagement would improve confidence."
        }
      ],
      "totalProfiles": 401,
      "scenarios": [
        "chronic_condition_intake"
      ]
    }
  ]
}

Start a job

Authorization: Bearer sk_live_…
X-Enterprise-Name: suki
Content-Type: application/json

{
  "patientPopulationIds": ["35e7f35d-a029-4639-91a6-fc4d5fdc668c"],
  "scenarioKeys": ["chronic_condition_intake"],
  "clinicianId": "9512b843-9b5c-4295-808c-08daa17d4705",
  "sessionsPerPopulation": 1,
  "maxTurns": 2
}

Success 202 (real sample job):

{
  "jobId": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "enterpriseName": "suki",
  "status": "queued",
  "links": {
    "self": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212",
    "export": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/export"
  }
}

Check job status

Authorization: Bearer sk_live_…
X-Enterprise-Name: suki

GET /api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212

{
  "jobId": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "enterpriseName": "suki",
  "status": "running",
  "progress": { "completed": 0, "failed": 0, "total": 1 },
  "config": {
    "clinicianId": "9512b843-9b5c-4295-808c-08daa17d4705",
    "clinicianName": "Medsurge Nurse",
    "patientPopulationIds": ["35e7f35d-a029-4639-91a6-fc4d5fdc668c"],
    "scenarioKeys": ["chronic_condition_intake"],
    "sessionsPerPopulation": 1,
    "maxTurns": 2
  },
  "error": null,
  "sessions": [{
    "sessionId": "38ee1fbe-0918-4211-b610-76449c628272",
    "patientPopulationId": "35e7f35d-a029-4639-91a6-fc4d5fdc668c",
    "scenarioKey": "chronic_condition_intake",
    "sessionIndex": 0,
    "status": "running",
    "error": null,
    "resultHref": null
  }],
  "export": {
    "ready": false,
    "href": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/export"
  },
  "createdAt": "2026-09-09T21:47:19.372Z",
  "updatedAt": "2026-09-09T21:47:25.104Z",
  "completedAt": null
}

Job status: queued → running → completed | completed_with_errors | failed
Session status: pending | running | completed | failed

When the job finishes, the same endpoint returns a terminal payload like:

{
  "jobId": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "enterpriseName": "suki",
  "status": "completed",
  "progress": { "completed": 1, "failed": 0, "total": 1 },
  "config": {
    "clinicianId": "9512b843-9b5c-4295-808c-08daa17d4705",
    "clinicianName": "Medsurge Nurse",
    "patientPopulationIds": ["35e7f35d-a029-4639-91a6-fc4d5fdc668c"],
    "scenarioKeys": ["chronic_condition_intake"],
    "sessionsPerPopulation": 1,
    "maxTurns": 2
  },
  "error": null,
  "sessions": [{
    "sessionId": "38ee1fbe-0918-4211-b610-76449c628272",
    "patientPopulationId": "35e7f35d-a029-4639-91a6-fc4d5fdc668c",
    "scenarioKey": "chronic_condition_intake",
    "sessionIndex": 0,
    "status": "completed",
    "error": null,
    "resultHref": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/sessions/38ee1fbe-0918-4211-b610-76449c628272"
  }],
  "export": {
    "ready": true,
    "href": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/export"
  },
  "createdAt": "2026-09-09T21:47:19.372Z",
  "updatedAt": "2026-09-09T21:48:06.711Z",
  "completedAt": "2026-09-09T21:48:05.213Z"
}

resultHref is set only when the session is completed. Follow that path with GET to load the session transcript. When export.ready is true, GET export.href for hrefs — not the transcripts.

Get a session

Authorization: Bearer sk_live_…
X-Enterprise-Name: suki

GET /api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/sessions/38ee1fbe-0918-4211-b610-76449c628272

Not ready — 409:

{ "error": "Session result is not ready yet", "code": "SESSION_NOT_READY", "status": "running" }

Failed — 200:

{
  "sessionId": "38ee1fbe-0918-4211-b610-76449c628272",
  "jobId": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "status": "failed",
  "error": "Engine timed out while generating the patient turn"
}

Completed — 200 (follow resultHref). Transcript uses nurse / patient. Patient turns may include citations (from source files). No evaluation block. job_id / session_id below match the sample job above; profile fields and dialogue illustrate the response shape:

{
  "schema_version": 1,
  "job_id": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "session_id": "38ee1fbe-0918-4211-b610-76449c628272",
  "session_index": 0,
  "scenario_key": "chronic_condition_intake",
  "profile_id": "e4a91c2b-6f3d-4a8e-9c17-2b0d5f8a1e33",
  "profile_name": "Maya Chen",
  "profile_summary": "31, first baby, scheduled C-section for breech. Under-describes incision pain until asked twice, redirects clinical questions toward the baby, and second-guesses every sensation after reading peer recovery stories online.",
  "transcript": [
    {
      "turn_index": 0,
      "sender": "nurse",
      "content": "Hi Maya, I'm Sarah — taking care of you after your C-section today. I want to start with how you are feeling right now. Where is the discomfort strongest when you shift between lying flat and sitting up?",
      "citations": []
    },
    {
      "turn_index": 1,
      "sender": "patient",
      "content": "Oh god — it's right across the bottom, where the incision is. When I push myself up it pulls so hard it feels like something's going to tear, even though I know it won't. Lying flat is fine; it's that in-between part that screams. And then I'm trying to figure out how to hold her without putting any weight on that spot.",
      "citations": [
        {
          "title": "First night after C-section — incision pain when sitting up",
          "url": "https://www.reddit.com/r/BabyBumps/comments/1c8xk2p/incision_pain_when_sitting_up_after_csection/"
        }
      ]
    }
  ]
}

Get the final results

Authorization: Bearer sk_live_…
X-Enterprise-Name: suki

GET /api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/export

This GET returns JSON. Poll until export.ready is true (job status completed or completed_with_errors), then call the same path as export.href. Use each resultHref with GET to load that session’s transcript. Export does not include analysis or transcripts.

If you call this GET too early — 409:

{ "error": "Export is not ready yet", "code": "EXPORT_NOT_READY", "status": "running" }

{
  "schema_version": 1,
  "jobId": "c2bb3f32-80a5-4966-82c2-bf7f3673a212",
  "status": "completed",
  "config": {
    "clinicianId": "9512b843-9b5c-4295-808c-08daa17d4705",
    "clinicianName": "Medsurge Nurse",
    "patientPopulationIds": ["35e7f35d-a029-4639-91a6-fc4d5fdc668c"],
    "scenarioKeys": ["chronic_condition_intake"],
    "sessionsPerPopulation": 1,
    "maxTurns": 2
  },
  "createdAt": "2026-09-09T21:47:19.372Z",
  "completedAt": "2026-09-09T21:48:05.213Z",
  "sessions": [
    {
      "sessionId": "38ee1fbe-0918-4211-b610-76449c628272",
      "status": "completed",
      "resultHref": "/api/v1/jobs/c2bb3f32-80a5-4966-82c2-bf7f3673a212/sessions/38ee1fbe-0918-4211-b610-76449c628272"
    }
  ]
}

sessions lists status plus resultHref (or an error stub if the session failed). Transcripts are only on the session GET.