Jio Arogya · Smithian Analysis
This site is access-restricted.
Please enter the password to continue.
Incorrect password. Try again.
← Home

Jio Arogya — Smithian Risk & Deal Dashboard

15 Joints · 18 Actors · 4 Rings · Load-bearing risks · Designed mitigations · Data streams
MEMBER & FAMILY TECHNOLOGY · AI RELIANCE HUMAN EXTERNAL
Ring 0Member & Family (2)
Ring 1Personal Agent · Clinical AI · AI Scientist (3)
Ring 2CHW · Supervisor · Nav Doctor · Care Entity · Allianz · Netmeds (6)
Ring 3Hospitals · BioPharma · Regulators · Trust · State Missions · Competitors (7)
Does the structure of Jio Arogya make selfish choices productive?
Smith's insight: systems work when self-interest and collective benefit point the same way, and break when they don't. The butcher feeds you because feeding you pays him. The system works not because people are good but because the structure makes good behaviour profitable.
Treating each stakeholder as a monolith misses the real dynamics. The CHW is not one person — she is herself, her supervisor, the ops CFO, three layers inside one "stakeholder" that do not always want the same thing. The AI is a fourth layer between the member and every human. And the member is not a solo actor — in India, health decisions are made by the family.
15 joints where rational self-interest pulls against alignment by default. At each joint, a specific designed value exchange — payment structure, indemnity, reputation signal, contractual term — converts the default conflict into productive behaviour. Every row carries a D-tag where it maps to one of the 8 named 90-day decisions (D1–D8). Every row names the data streams that the joint produces, consumes, or audits.

What this dashboard is

This is an operational risk-and-deal tracker for the Jio Arogya system. It maps 15 joints — the specific points in the workflow where two or more rational actors' self-interest pulls against alignment by default — and the designed value exchange at each joint that converts the default conflict into productive behaviour.

The dashboard is the working surface for the team preparing mechanism-design interventions. The full analytical narrative is in the companion document Smithian-Actor-Chains_Jio-Arogya.md.

How to read the table

  • # — Joint number (1–15), matching the section numbering in the Smithian doc.
  • Joint / Actor Chain — The named tension. Click any row to open a detail panel with the full six-layer Smithian analysis (individual, layer above, org objective, AI intersection, escalation boundary, designed value exchange).
  • Ring — Which ring of the system the joint sits in. Ring 0 Member & Family. R1 · AI Technology quasi-actors. R2 · Rel Reliance-controlled human actors. R3 · Ext External counterparties.
  • Key Risk — What happens if nobody designs the value exchange and rational actors follow their default incentives.
  • Data Streams — The data the joint produces, consumes, or audits. From the Jio_Care_Dataset_Inventory.md seven-category taxonomy.
  • Proposed Deal Mitigation — The specific mechanism (payment, indemnity, contract term, governance structure) that converts the conflict into alignment. From the Smithian doc.

What the badges mean

Load-Bearing
The mechanism is load-bearing and missing from the plan. Must be designed before the system ships. (3 joints)
Implied
The plan implies the mechanism but does not name it explicitly. Needs to be surfaced and committed in writing. (7 joints)
In Plan
The mechanism is already in the plan, either fully or partially. Validate and lock. (5 joints)
D1–D8
Maps to one of the 8 named 90-day decisions that cannot wait until the business plan closes in July. Full register in JioCare - Strategic Decision Layers.md.

The seven data-collection families

Every joint produces, consumes, or audits data. The "Data Streams" column and the detail panel's "Upstream Data Categories" both reference these seven families from Jio_Care_Dataset_Inventory.md. Click any N.Category badge in the detail panel to jump to the Glossary.

#FamilyWhat it covers
1OmicsWGS, SNP, PRS, pharmacogenomics, methylation, proteomics, metabolomics, microbiome, transcriptomics — Unified Blood Draw + Gut Health Kit
2Biomarkers / LabMetabolic baseline, lipid, liver, inflammatory, advanced 100+ biomarkers, kidney — fasting blood draw + urine
3ImagingFull-body MRI, CCTA, calcium score, DEXA, carotid US, retinal imaging — heavy iron + point-of-care
4Clinical / AssessmentVO₂ max, grip strength, cognitive assessment, foot/neuropathy, body composition, BP — mix of clinic, home, and digital
5Continuous / WearableCGM, Oura/WHOOP, activity, smart scale, home BP, food logging, stress/biofeedback — passive + active
6Phenotypic / ContextualLifestyle, behavioural, smoking cessation, socioeconomic, intervention protocol, medication titration, outcome metrics
7Consent / GovernanceDPDP consent ledger, ABDM interoperability, research consent artefacts, audit trail

Suggested workflow

  1. Start with Load-Bearing Missing. Filter to "Load-Bearing Missing" — these 3 joints (Doctor-AI liability, Netmeds pharmacy choice, Member-Family co-decision) are where the system breaks if no mechanism is designed.
  2. Then 90-Day Decisions. Filter to D-tags — these 6 joints map to decisions that must be made in the next 90 days.
  3. Review Implied joints. These 7 are the quiet risks — the plan assumes them but hasn't committed in writing.
  4. Lock In Plan joints. Validate the 5 joints where mechanisms already exist.
  5. Click any row to open the detail panel with the full Smithian six-layer analysis for that actor.
  6. Read mitigations in the table — the proposed mechanism for each joint. Click a row for the full six-layer analysis.

The document family

This dashboard sits inside a family of analytical documents. Each takes a different lens to the same system. Read them in the order that matches your question.

DocumentWhat it doesWhen to read it
Proposal Memo_200426.md The master business plan. Six components: daily engagement, navigation, intervention-budget engine, biobank, Research Trust, insurance interface. Four-tier hospital architecture. CHW cadre at 1:1000. Privacy-by-architecture. Data Dividend as civilizational contract. Start here if you haven't read the plan. This is the system being analysed.
260421_Strategic_Memo.md The two-page strategic pre-read. Frames the opportunity (cardiometabolic claims, biobank economics, AI clinical quality), the six-component operating system, and why only Reliance can build it. Read before the Proposal Memo if you want the 5-minute framing before the 45-minute detail.
260421_Data_Strategy.md The data thesis. Longitudinal linked data as the compounding asset. Defines the seven data-collection families (Omics, Biomarkers, Imaging, Clinical, Continuous, Phenotypic, Consent/Governance) that appear in every joint's "Data Streams" column and detail panel. Pharma demand sizing. 43× premium for depth (23andMe vs deCODE). Data Dividend mechanics. Read when you need to understand the data streams column in this dashboard, the seven-category taxonomy, or the Research Trust licensing model. This is the canonical source for what each data family contains and how it is collected.
JioCare - Strategic Decision Layers.md The 90-day sprint decision surface. Eight domains (care model, data/tech, business model, operating model, capital, org design, regulatory, partnership) each broken into the decisions the team must make before July. Read when you need the D1–D8 decisions in full operational context, or when the dashboard's decision tags need background.
Smithian-Actor-Chains_Jio-Arogya.md The primary source for this dashboard. 18 actors in 4 rings, 15 joints, 6 questions at each joint (individual, layer above, org objective, AI intersection, escalation boundary, value exchange). This is where the detail panel data comes from. Read the full document when a joint's detail panel raises a question the summary doesn't answer. Each joint has a §-reference (e.g. §2.1, §3.2) pointing to the relevant section.
Verbal-Brief.md The verbal brief and Q-index. Anticipatory Q&A mapped to the plan's weak points. The Q-index maps to the dashboard's load-bearing joints.
260421_Principles_Decisions_Op_Model.md Principles, key decisions, and the operating model. Covers org design, reporting lines, the three-entity structure (Care Entity + Allianz JV + Research Trust), and the single-owner question. Read when Joints 7, 8, or D1/D3/D7 are under discussion — the transfer-pricing, AI ownership, and org-design decisions live here.

Entities & Actors

Care Entity
Jio Arogya Care Services — the care delivery org. Runs the CHW cadre, navigation doctors, member experience. One of three P&Ls (alongside Allianz JV and Research Trust).
Allianz JV
Jio Allianz — the insurance joint venture. The wrapper, not the business. Provides the captive payer, regulatory license, and actuarial discipline.
Research Trust
Separate legal entity (proposed Section 8 company) holding de-identified longitudinal data. Produces pharma licensing revenue and the Data Dividend. Independence from Reliance is load-bearing for credibility.
CHW
Community Health Worker (Arogya Saathi). Salaried, 1:1000 member ratio. The first human the member sees after the AI decides a human is needed.
Personal Agent
On-device AI. Holds all PII locally, orchestrates downstream layers. Daily touchpoint with the member. Most powerful actor in the system.
Clinical AI
Stateless inference service on Jio Brain. Different team from the Agent. Closer to medical literature. Authoritative at clinical decision points.
AI Scientist
Autonomous research system on the Trust's de-identified dataset. Compresses multi-year research cycles. Year 4+ capability.
Navigation Doctor
MBBS (possibly PG), NMC-registered. Clinical authority in the access stack. Authorises AI-recommended actions. Carries the liability the AI cannot.
ASHA
Accredited Social Health Activist. Government frontline health worker under the National Health Mission. ~1 per 1,000 rural; 1 per 2,500 urban. ₹3,500 central fixed incentive + state top-ups.
Netmeds
Inside Reliance Retail. Pharmacy dispensing rail. Most regular data feed in the system. Also the single largest regulatory tying-risk surface.

Regulatory Bodies

IRDAI
Insurance Regulatory and Development Authority of India. Regulates insurance products, bundling, commission structures, and the Allianz JV.
NMC
National Medical Commission (replaced MCI). Regulates medical practice, doctor registration, telemedicine guidelines. The body that can suspend a navigation doctor's license.
ICMR
Indian Council of Medical Research. Governs biobank ethics, research consent, biomedical research guidelines. Approves the Research Trust's research activity.
DPDP Authority
Digital Personal Data Protection Authority (under the DPDP Act 2023). Governs consent architecture, data processing basis, member rights, and the Data Dividend's legal frame.
CDSCO
Central Drugs Standard Control Organisation. Could classify the AI triage platform as a medical device (Class B or C), triggering registration and clinical-trial requirements.
NABH
National Accreditation Board for Hospitals & Healthcare Providers. Quality accreditation framework. Used in the capitation scorecard as the Indian analogue to HEDIS.

Data Collection Families (7 categories)

From Jio_Care_Dataset_Inventory.md and 260421_Data_Strategy.md. Each joint in the dashboard references these by number. The N.Name badges in the detail panel link here.

1. Omics
Whole-genome sequencing, SNP arrays, polygenic risk scores, pharmacogenomics, methylation, proteomics, metabolomics, microbiome, transcriptomics. Collected via Unified Blood Draw + Gut Health Kit at enrolment.
2. Biomarkers / Lab
Metabolic baseline, lipid panel, liver function, inflammatory markers, advanced 100+ biomarker panel, kidney function. Collected via fasting blood draw + urine sample.
3. Imaging
Full-body MRI, coronary CT angiography (CCTA), calcium score, DEXA bone density, carotid ultrasound, retinal imaging. Mix of heavy-iron (hospital) and point-of-care devices.
4. Clinical / Assessment
VO₂ max, grip strength, cognitive assessment, foot/neuropathy screening, body composition, blood pressure. Mix of clinic visits, home-based, and digital capture.
5. Continuous / Wearable
Continuous glucose monitor (CGM), Oura/WHOOP ring, activity trackers, smart scale, home BP cuff, food logging, stress/biofeedback. Passive + active daily capture.
6. Phenotypic / Contextual
Lifestyle surveys, behavioural assessments, smoking cessation status, socioeconomic indicators, intervention protocol logs, medication titration records, clinical outcome metrics.
7. Consent / Governance
DPDP Act consent ledger, ABDM health-claims interoperability artefacts, research consent records, audit trail. The legal and compliance backbone of the data asset.

Technical & Structural Terms

Access Rule
No CHW self-dispatch; ordered layer escalation. The structural feature that keeps the CHW cadre inside the intervention budget.
Data Dividend
The civilizational contract: members share health data; the system returns financial value from pharma licensing. Weighted by tenure and participation, not disclosure completeness.
PMPM
Per Member Per Month. The unit cost metric for care delivery. The Care Entity → Allianz transfer price is expressed as PMPM.
HCX
Health Claims Exchange. India's ABDM-linked claims interoperability standard. Used for Tier 2 hospital discharge handshake.
HTA-In
Health Technology Assessment in India. The body that evaluates clinical AI safety and can serve as the independent auditor for the weight-change review board.
Section 8
Section 8 of the Companies Act 2013 — the Indian legal structure for a non-profit company with governance requirements. Proposed vehicle for the Research Trust.
SaMD
Software as a Medical Device. FDA classification framework increasingly referenced by Indian regulators for clinical AI.
Capitation
Fixed payment per member per year to a hospital, regardless of services rendered. Inverts overtreatment into undertreatment risk.
Risk Corridor
A loss-ratio band around novel interventions: Allianz guarantees the band; Care Entity absorbs excess. De-risks the actuary's catalogue-freeze instinct.
Jamnagar Pilot
The captive-cohort launch: 10K Reliance-employee families (~40K members), full architecture from Month 7. The short-term moat before the data asset compounds.
Filter: All (15) Ring 0 · Member (1) Ring 1 · AI (3) Ring 2 · Reliance (6) Ring 3 · External (5) Load-Bearing Missing (3) Implied Not Named (7) Already In Plan (5) 90-Day Decisions (D1–D8)
# Joint / Actor Chain Ring Key Risk (natural default) Data Streams Proposed Deal Mitigation

Natural Synergy — Preserve, Don't Redesign (§6)

  • Personal Agent · Member — properly specified, the Agent's job is to help; the member wants help. Design work is in weight-setting governance, not in the relationship.
  • Netmeds · Member — naturally aligned (volume vs medicine). Risk is tying, not incentive.
  • Research Trust · Data-Dividend Member — Dividend makes the member a co-owner of the asset. Civilizational alignment. Easy to break if Dividend is token/late/opaque.
  • CHW · Assigned Families — social alignment already in place. System must not break it via rotation.
  • Allianz · Long-Tenure Member — statistically cheaper to underwrite. Premium must reward tenure, not penalise claims.
  • Family · CHW — where the CHW is accepted into the household, the informal network becomes an asset, not a risk.
Source: Smithian-Actor-Chains_Jio-Arogya.md + Jio_Care_Dataset_Inventory.md · 4 rings · 15 joints · 7 data categories
3 load-bearing & missing · 7 implied not named · 5 already in plan · 8 named 90-day decisions
JOINT #0
—