RailTel · APMSIDC · T. No. 1.2/APMSIDC/2026-27 · presentation scored /20 (RFP §3, p. 53)
रेलटेलRAILTELA Navratna CPSE · Govt. of India
Government of Andhra Pradesh · APMSIDC
SANJEEVANI
Approach & Methodology
Statewide Integrated Digital Care Coordination and Healthcare Delivery System, Andhra Pradesh
Tender T. No. 1.2/APMSIDC/2026-27 · scored out of 20 marks (RFP §3, p. 53)
11,736facilities · 28 districts
5 crore+citizens
105,492health functionaries
60months from effective date
Presented by RailTel Corporation of India Ltd A Navratna CPSE, Ministry of Railways — with Suja HealthTech
A
Technical Capability & Relevant Experience of Bidder
RailTel — a Navratna operator whose own assets answer this project's largest delivery risks; credentials submitted separately with the proposal.
Scored via documents
100+years of collective experience behind Sanjeevani
25+RailTel — Navratna CPSE, est. 2000 · national networks, NCII operations, the Railways health platform
27+Suja HealthTech — healthcare-only since 1999 · 2,500+ installations across India & the US
30+RailTel Consultants & advisors — Dr Viduthalai Virumbi · Dr Sankar Reddy · Mr Roop Kumar · Mr Prabhudoss and the Sanjeevani advisory team
Domains of experienceHealthTechPublic HealthOne HealthPlanetary HealthWar RoomMass GatheringsOutbreak / Pandemic ResponseDisaster ResponseClinical GovernancePopulation GovernanceOperational ResearchDevelopment Research
11,000+points of presence (6,117 railway PoPs)
63,000+route-km fibre on railway RoW
21,000+km of city access network
Tier-IIIMeitY DCs · 2 edge DCs · 24×7 SOC
~695hospitals & health units on RailTel cloud
~80 lakhRailways health beneficiaries
NavratnaAug 2024 · debt-free · profit-making
NCIINational Critical Information Infrastructure
The Indian Railways hospital & health platform — the same problem shape as Sanjeevani — already runs on RailTel
infrastructure. The solution partner is Suja HealthTech — 27 years healthcare-only since 1999, 600+ healthcare-IT
professionals, 2,500+ installations, ONC-certified through every stage — bringing CareAxes HMIS, Gyneye and production AI.
Full detail
A
Corporate profile — the numbers behind the confidence
Central NOC at New Delhi with four regional NOCs — Mumbai, Delhi, Kolkata, Secunderabad; 4 regions and 21 territories covering every state.
Unified licence — NLD, ISP & ILD, IP-1; no network elements from border-sharing countries.
Why it matters here: connectivity at 10,032 Village Health Centres is the risk most likely to stop a facility going live — a bidder operating national fibre answers it with assets, not subcontracts.
A
RailTel in health — the anchor citation
Scored via documents
Our anchor citation is a health system, not a platform: ~695 hospitals and health units serving ~80 lakh beneficiaries on RailTel cloud — registration to discharge, telemedicine, unified identity — operated on our own infrastructure.
All four national digital-health building blocks (ABHA, HFR, HPR, HIE-CM) live in a working system.
In the live demonstration, a record created in the Railways system is fetched into Sanjeevani on the citizen's consent — interoperability shown, not certified.
Sovereign posture: data and model processing within India; open-source model on-premise in the State Data Centre; IP and weights vest with the Authority.
A
Suja HealthTech — the solution partner, 27 years healthcare-only
Scored via documents
Founded in 1999 in New Jersey and healthcare-only ever since — never merged, never acquired, never a
discontinued product. In-house product, implementation, training and L1/L2/L3 support: one accountable partner, not a
consortium of subcontractors.
27years in healthcare IT — since 1999
600+healthcare-IT professionals
2delivery centres — New Jersey (HQ · product engineering) & India (implementation · support · R&D)
2,500+installations supported
153K+providers & system users
65M+annual EDI transactions
99%first-attempt clean claims
$160Msaved for CMS — population health across 1,652 GPTN sites, 5,287 providers
A continuous track record
1999 — Founded in New Jerseyhealthcare technology from day one
▼
2006 — First CCHIT certificationthe first of five certified editions
▼
2011–15 — Meaningful Use I, II, IIIcertified through every ONC stage
▼
GPTN — $160M saved for CMSpopulation health across 1,652 sites
▼
Today — CareAxes, AI-enabledone unified platform, India & US
Trusted by providers in India & the US
Fernandez Hospital, Hyderabadenterprise CareAxes · multi-site · fully paperlessA4 Hospitals, Chennaimulti-specialty enterprise deploymentChennai Urology & Robotics InstituteCareAxes enterprise deploymentNew Hope Foundation, USMedicsCloud EMR · revenue cycleNewbridge Medical Center, USMedicsCloud EMR · revenue cycleWoodbridge Medical Center, USMedicsCloud EMR · revenue cycle
Production AI — built inside healthcare workflows
MedicsScribeAIclinical documentationMedicsSynopsiscontext-aware summariesMedicsHCCcoding supportRPMremote patient monitoringAI fax routingdocument automationVirtual Group Notesgroup documentationSUJA Agentagentic multi-step workflowsGYNEYEAI / vision models for colposcopy
The India footprint — CareAxes
The India-facing HMIS where Suja's healthcare experience is localised — hospital
workflows, multi-branch integration, specialty modules and configurable deployment · dashboards, analytics, quality measures and a
research-friendly data architecture · HIPAA-aligned, HL7 FHIR, cloud or on-premise · a women's-health runway where birthing and
fertility become the foundation for cervical screening, menopause care and longitudinal women's health.
Legacy absorption & one-time cutover with 14-day reconciliation
Data lake registries scored monthly — scores gate the AI layer
The flow on the Software page runs on this platform: the family-and-street register is the denominator, the four blocks are its
workloads, and the AI engine's insights ride the same ABHA & geotag spine
C
Project Plan with Timelines
The five-year term as an interactive schedule — drill from the full term to any quarter, month and week, filtered by delivery stream.
3 marks
Full detail
C
Master phases and the arithmetic that governs the plan
3 marks
Phase
Duration
Gate
Position
Assessment, integration & testing
3 months
Survey report · signed specification · approved sizing · Centre built · full complement on board
Self-funded; no payment in this phase
Effective date
—
On completion of the assessment period
The operating clock begins
Implementation
6 months
Total coverage across all 28 districts · completion certificate
Capital release at 3 months and on the certificate
Operation & management
54 months
24×7 service under agreed service levels
Fee drawn over twenty quarters
Exit
Final 6 months
Escrow · IP vesting · asset & knowledge handover
Performance security released after handover
The operating clock begins at the effective date, and the six-month implementation falls inside it. Steady-state operation is therefore 54 months, not 60 — every staffing curve, training cycle, release train and indicator ramp in this plan is built on that arithmetic.
D
Proposed Team
One accountable leadership spine — the Sanjeevani Authority above, the Programme Director at the centre, three divisions below, and the 275-seat Centre underneath.
Monitors programme operations — engineering, integration, security, QA and corporate functions — and coordinates with the
authorities through the Programme Director.
Programme DirectorDr Viduthalai Virumbi
Single point of accountability to the Sanjeevani Authority — supervising the programme from the self-funded assessment
through implementation and the five-year operation. Chairs the delivery reviews of all three divisions, leads the monthly
coordination committee, and stands available to the Authority throughout the term.
Advisory Team
Public-health, clinical, digital-health and governance advisers — standing counsel to the Programme Director and the Authority's reviews.
Mr RajinikanthPhygital Division
Software Division — platform, citizen & field apps, integration, data lake & AI
Hardware Division — Centre estate, facility devices, connectivity, hosting & DR
Dr Sankar ReddyHealthcare Division
CPHC — comprehensive primary health care: service delivery & clinical operations
Care pathways, STG adherence, clinical QA and the consultation cabins' medical practice
Across three shifts — capacity 275 seats; peak concurrent ≈ 135
Headroom absorbs shift handover overlap, the daytime weighting of outreach, campaign and outbreak surge, training rotations and Authority positions. Additionally stationed for the full duration: installation & field team, medical service provider team, equipment analyst & support and data analysis & reporting.
D
The 37 named roles
3 marks
Role
No.
Profile
Programme Director
1
MPH/MHA/MBA-equivalent · 15 yrs · supervising through implementation and operation, available to the Authority
District Coordinator
28
MPH/MHA/MBA/Tech · 6 yrs · one per district
Platform Lead
1
B/M.Tech CS-IT · 12 yrs
Public Health Specialist
2
MD/MPH/Community Medicine · 6 yrs
Data & AI Lead
1
B/M.Tech / MSc Data Science · 8 yrs
Cybersecurity Lead
1
Recognised infosec certification · 8 yrs
Centre Ops · Field Ops · Telehealth Clinical leads
3
Named in the annexures with CVs in the prescribed form
Where this heading's marks actually sit: named individuals against the nine key roles, CVs in the prescribed form, and qualification fit — checked against the submitted annexures. The content is in this section; the evidence rides with the proposal.
D
Governance, escalation and mobilisation
3 marks
Escalation — and what bypasses it
L1 Facility/field → district coordinator — same working day; resolve in 48 h.
L2 District → field-ops / platform lead — 24 h; resolve in 5 working days.
L3 State workstream → workstream lead — 24 h; tabled at coordination committee.
L4 Programme → Programme Director — immediate; steering committee.
Four classes bypass the ladder and notify at programme level immediately: any security or data incident · any clinical-safety event on advisory guidance · any event that would breach a service level · anything that would delay the coverage milestone.
Mobilisation — front-loaded into the assessment period
Wk 1–2 PD & leads on; survey teams to all 28 districts; Centre lease executed; HR plan submitted.
Wk 3–6 Survey at scale; engineering/integration/data teams; Centre drawings; recruitment at scale; staging live.
Wk 7–10 Registers complete; FRS; sizing submitted; fit-out under way.
Q1 Trainers deployed ahead of each wave; Centre soft launch.
Fit-out, console, wall and cabin suppliers pre-qualified with long-lead items against letters of intent — procurement does not consume the assessment period.
E
Proposed Solutions, Component Design & Readiness
A seven-layer solution architecture on a production estate — channels to Kubernetes, with the State Health Data Lake and the external ecosystem integrated, not replaced.
5 marks
Three differentiators — what no other bid brings
Readiness, stated plainly
Production: platform · channels · contact centre · command wall · integration · lake · BI · Telugu speech recognition & voice analytics. UAT: verification & decision-intelligence layer. Adapted: sovereign agent layer. In build, named first: cervical-model regulatory position · health score · Centre fit-out.
The ICCC video wall — common operational picture, live from the working prototype (28 districts · 11,736 spokes · CEP engine)
Full detail
E
Differentiator 1 — GIS & address hierarchies under every report
5 marks
Two hierarchies, one geography — converging on the village
Health & administrative
StateGovt of Andhra Pradesh
▼
28 districtsdistrict hospitals & DM&HO
▼
MandalsCHCs & area hospitals
▼
PHC areasprimary health centres
▼
HSC areashealth & wellness sub-centres
▼
Village / wardthe household register
Political constituency (LGD)
State
▼
25 Parliamentary constituenciesMP
▼
175 Assembly constituenciesMLA
▼
Village / wardLGD ↔ AC ↔ PC join — CEO-AP roll
What this buys the State
Every record geo-anchored at capture — facility and household, down to the village register
The video wall drills District → Mandal → PHC area → HSC area → Village on the live KPI roll-up
Every village mapped to its MLA & MP constituency — the same report renders by administrative, health-facility or constituency hierarchy
Choropleths & heat maps on any indicator at any level — outbreaks, stock, coverage, campaign progress
Blackspot mitigation and wave planning run on the same geography — one address spine, no re-mapping
E
Differentiator 2 — ABHA as the fiduciary ID
5 marks
One citizen, one spine — every identifier linked to ABHA
AadhaarABHA seeding & e-KYC
NTR Vaidya Sevascheme member ID · sanctions & balance
GSWS household IDfamily folder · rice card
ABHA14-digit Ayushman Bharat Health Account
Nikshay IDTB programme
RCH IDmother & child tracking
Facility MRNs11,736 local registers absorbed
Why fiduciary, and why it matters
One longitudinal record — programme IDs stop being silos; deduplication happens once, at the identity layer
Consent-bound access through ABDM HIE-CM — purpose-limited, auditable, revocable
Claims ride HCX keyed to ABHA — NTR Vaidya Seva sanction, balance and settlement against the same identity
Portability — the record follows the citizen across all 11,736 facilities and every scheme
Privacy engineered — masking, role-based views, fiduciary duty on every custodian of the linked IDs
E
Differentiator 3 — configured for the Govt of AP
5 marks
The Sanjeevani digital backbone — configuration, not a fork
1statewide instance — every facility tier configured, none forked
AP's own registers preloaded — LGD codes, 28 districts, facility master, NTR Vaidya Seva, GSWS family folders
Telugu voice-first across citizen and field apps — production speech recognition tuned on the State corpus
State integrations as first-class — 104/108 continuity, IHIP, Nikshay, U-WIN, RCH, GoI HMIS, CM dashboards
Governed as the State's asset
Change envelope & architecture review (S2) — the Authority controls what changes and when
Release train — minor monthly, major quarterly, every release gated and reversible
Exit engineered from day one — documentation & escrow verified, incoming-agency training, handover of capability, not just code
Configuration survives upgrades — a fork would strand the State; configuration keeps AP on the product's mainline
E
The solution architecture — seven layers, one record
5 marks
One journey through every layer
1Spoke or CHW initiates→2Gateway authenticates · consent checked→3Encounter captured in the specialty EMR→4AI assists — the clinician confirms→5The hub consults & coordinates→6Written to the record, the lake & ABDM/PHR→7Dashboards & the ICCC wall→8Closed-loop referral & follow-up
1Channels & experiencepresentation tier · Telugu & English · offline-capable
2API gateway, identity & consentauthenticated by default · every call scoped to facility & district
API gatewayOAuth 2.0 · throttling · versioning · FHIR façadeIdentity & tokensJWT · SAML 2.0 SSO with the State IdP · 2FA/MFAAuthorisation modelRBAC · role catalogue · menu & button permissionsConsent check — HIE-CMpurpose · scope · audit before any record is served
3Application servicesdomain microservices · API-first · event-driven between services
Facility clinical core — in production today · the system of record for the encounter inside the facility
Inside a service: controller → domain service interface → infrastructure service → Unit-of-Work → PostgreSQL · between services: REST/FHIR + the event bus.
6Data & interoperability — the State Health Data Lakehot / warm / cold tiering · governed end to end
1Sources — clinical core · CHW · contact centre · legacy · GoI/State APIs · analysers→2Ingest — streaming/CDC · batch · API pull, with audit & reject queue→3Raw landing — immutable, encrypted, replayable→4Standardise to FHIR R4 — terminology · ABHA identity resolution→5Curated — programme marts · denominators to habitation · ML features→6Serve — BI · agents & models · PHR/ABDM · RTGS-Lens
PostgreSQL 16 — system of recordtransactional store — never serves the lake, so analysis cannot slow a consultationTime-series storevitals & device streams, queue depth, TAT — aggregates, compression, retentionVector store for AIsemantic search over protocols · GenAI retrieval · patient similarityABDM building blocksABHA · HFR/HPR · HIE-CM · UHI · PHR — source and consumerGovernance railcatalogue & lineage · PII masked · consent registry · quality rules · retention · full audit · all data in India
7Platform, scaling & observabilityenterprise CNCF Kubernetes — SDC + DR · no single point of failure
Risk position: third-party API readiness is assessed during architecture & implementation planning; delays outside the Agency's control sit outside SLA and timelines, the Agency bears integration cost, and the Authority coordinates third-party changes.
E
Readiness, component by component
5 marks
Component
Readiness
Assessment-period work
Core platform & care pathways
Production
Configuration to State workflows and protocols
Citizen & field applications
Production
Consolidation, Telugu & offline extension, store publication
Contact centre & grievance
Production
Number continuity, routing, disposition taxonomy
Command & incident platform
Production
Event rules, geographic layers, wall mosaic
Integration layer · data lake · BI
Production
Connector build per target; ingestion against disclosed volumes
Agent layer & sovereign model
Production, adapted
On-premise in the State DC; role-based prompting; protocol validation
Telugu speech recognition & voice analytics
Production
Dialect-coverage validation & tuning on the State corpus
Cervical image model — regulatory
To be established
Registration pathway; deployed as decision support, clinician in loop
Health score algorithm
To be defined
With experts nominated by the Authority; fairness & audit designed in
Verification & decision-intelligence layer
UAT
Deploys after the survey — you cannot enumerate a population you have not surveyed
A readiness claim that is not true is discovered at the live demonstration. Five components are not in production; all five are named here before they are asked about.
E
The Centre — five missions on one floor plate
5 marks
Five operating missions
Outbound care coordination — ~90% of call volume; follow-up and defaulter management.
Clinical consultation — 40 video cabins, the clinical heart of the hub.
Command & control — the common operational picture, CEP, campaign and incident operations.
Programme intelligence — analytics floor feeding State reviews.
Engineered, and benchmarked
36,000 sq ft on one floor plate; zoned floor; acoustics, power and safety engineered; three-shift resilience.
Benchmarked against command centres actually built — and honest about the evidence: on command centres the published evidence is mixed, which is why every Centre function is tied to a measured care loop rather than to a wall of screens.
Facility zoning — 36,000 square feet on one floor plate
Concentric by design — shared visual picture at the front, high-density coordination floor facing it, enclosed cabins
and pods on the flanks, command and support around the perimeter. Every block earns its place — the operations room is largest
because outreach is the core mission; cabins are enclosed for eye-line, lighting and privacy; pods sit beside the clusters they
supervise. Capacity exceeds concurrency — 154 seats in the war room, ≈ 82 occupied at day peak: the headroom is surge capacity
for campaigns and outbreak response.
Centre detail
ICCC — key points
24×7 in three shifts — 275-seat roster · ~90% outbound coordination
CEP engine correlates events from every module — beds, stock, labs, teleconsult load
Video wall drills District → Mandal → PHC area → HSC area → Village — each village mapped to its MLA & MP constituency
Emergency activation — a citizen's 108 self-report, disaster, mass-casualty and campaign surges each open an incident with its playbook
104 continuity and grievance intake with 48-hour response
QA on the floor — call scoring, triage adherence, monthly audit
CI/CD deployment — three environments, two approval gates
5 marks
Access narrows as code moves toward the citizen
Environment 1DevelopmentData & platform code — accessible to
Suja HealthTechRailTelGovt of AP
Where the product is built and iterated — every feature, fix and configuration starts here, under CI with automated build & test on each commit.
→Suja pushes code to Staging only on RailTel approval
Environment 2StagingData & platform code — accessible to
RailTelGovt of AP
Where releases are verified and hardened — a production-mirror where RailTel and the State test each gated release before it goes anywhere near the citizen.
→RailTel pushes code to Production only on Govt of AP approval
Environment 3ProductionData & platform code — accessible to
Govt of AP
Production data is accessed by Govt of AP alone. The statewide instance serving all 11,736 facilities — every release that reaches it carries two recorded approvals.
🔐Platform code in escrow — from day 1 to exit. The full platform codebase sits in escrow throughout the
term and is verified with every release, so the State's continuity never depends on any vendor.
The access matrix, stated plainly
Party
Development
Staging
Production
Suja HealthTech
Data & platform code
Code push only — on RailTel approval
No access
RailTel
Data & platform code
Data & platform code · approves Suja's pushes
Code push only — on Govt of AP approval
Govt of AP
Data & platform code
Data & platform code
Data & platform code — sole access to production data
Why the pipeline is shaped this way
Every promotion is an approval on record — nothing reaches Staging without RailTel's sign-off, nothing reaches Production without the State's
Sovereignty by construction — production data never leaves the Govt of AP's custody; no vendor, prime or partner, can touch it
Separation of duty — the party that builds (Suja) is not the party that verifies (RailTel), and neither is the party that owns (Govt of AP)
Exit engineered from day one — the escrowed platform code, verified release by release, means handover is a formality, not a negotiation
E
The Centre, benchmarked against what has actually been built
5 marks
There is no published purpose-built telemedicine command centre to benchmark against. The strongest precedents are
hospital capacity command centres, for the analytics floor, and Indian integrated command centres, for the wall, the zoning and the
control-room discipline.
Facility
Floor area
Shared display
What this design takes from it
A leading US academic capacity command centre
5,200–5,500 sq ft
22 screens · 14 analytic tiles
Co-located pods and the wall-of-tiles model; capacity released without new headcount
A commercial hospital command-centre platform
≈ 500 hospitals
Predictive tiles
Census and risk-forecast tiles as the model for agent-driven dashboards
A large Indian State integrated command centre
6.42 lakh sq ft
Multi-agency wall
Multi-tenant event correlation and control-room layout discipline
Indian smart-city command centres, typical
1,500–16,000 sq ft
3×3 to 6×4, 4K-native
Video-wall grid norms and the maturity framework for such centres
The national telemedicine service
Distributed
—
163 million consultations — this State the highest adopter
The key adaptation is scale. A hospital command centre coordinates one building; this Centre coordinates a
State. We keep the wall-of-tiles and the co-located pods, scale the floor to a 275-seat contact-and-care operation, and add forty
enclosed clinical consultation cabins that the hospital model does not have.
E
The shared visual picture — why the numbers are what they are
5 marks
7.2 × 2.025 mprimary video wall · P1.5 LED · 4,800 × 1,350
6×image height to the farthest viewer — 12.1 m, beyond the 10 m back row
26°top-of-wall vertical angle at the 3.5 m front setback — inside the ISO 30° limit
P1.5places front consoles in the comfortable band while overview content reads at the back
The method, not taste
AVIXA viewing-distance method sizes the image against the farthest seat
ISO 11064 sightlines cap the vertical viewing angle at 30°
Image height of 2.025 m puts the farthest viewer at six times image height
The wall carries the common operational picture — tiles, not decoration: every tile tied to a measured care loop
E
Forty consultation cabins — the clinical heart of the hub
5 marks
Each cabin is a single-clinician booth on the hub side; the patient is remote at the spoke, assisted by the
community health officer or activist. Sized larger than an open desk because of acoustic separation and camera geometry.
Element
Specification
Footprint
2.4 × 2.9 m — ≈ 7 m² per cabin
Display & camera
32-inch minimum with speakers; HD/4K 13 MP camera at seated eye level, 0.8–1.0 m from the clinician
Lighting
Diffuse front lighting, 400–500 lux vertical at the face, 3000–3500 K — never backlit, never a window behind
Background
Powder-blue matte acoustic panel; white, glossy and dark surfaces avoided
Acoustics
NC 30–35 ambient; partitions at STC 45–50; acoustic ceiling and resilient flooring
Privacy & supervision
Lockable door with a narrow glass vision panel
Occupancy
One hub clinician, with an optional scribe or specialist for joint consultations
4 × 10specialty clusters — general primary care · maternal & child health · NCD & mental health · specialist second opinion
1 : 10supervisor per cluster, built into the geometry
1.5 mservice corridor along each cluster
Podseach cluster adjacent to its clinical command pod
E
The Centre, engineered for continuous operation
5 marks
Element
Provision
Power
Tier-III N+1 UPS, 10–15 min autonomy, bridging to an N+1 generator — concurrently maintainable, so the floor never goes dark for service
Cooling
N+1 precision cooling for the server & hub zone; comfort conditioning with fresh air for the floor
Flooring
Raised access floor 300–600 mm in server and wall zones; anti-static tiles
Fire & safety
Aspirating smoke detection with clean-agent suppression in the server room; detectors, extinguishers, marked egress
Network
Redundant SD-WAN to all 11,736 spokes; managed switches, next-gen firewalls, structured cabling for all concurrent users
Security
Standard-aligned infosec; access control & CCTV; end-to-end encrypted consultations with audit logging
Ergonomics
Control-centre-standard consoles for the 5th–95th percentile; adjustable desks, acoustic partitions, glass supervision lines
99.982%availability at the Tier-III edge
1 : 10supervisor to agents, every shift
3shifts — continuous, throughout the year
11,736spokes on redundant SD-WAN
F
Detailed Work Plan incl. Resource Mobilisation
Three delivery streams — software, hardware and human resources — carry fourteen strategies across the five-year term; from sign-off, the digital backbone deploys to every facility.
2 marks
Three delivery streams — fourteen strategies, one term
🚀From sign-off, the Sanjeevani digital backbone deploys to every facility.
Six software, four hardware and four human-resources strategies carry the five-year term — the full chart sits one click below.
Full detail
F
The fourteen delivery strategies across the five-year term
2 marks
M−3 · assessmentM6 · full coverageM30M60 · exit
S1Requirements & assessment
S2Specification, design & UX study
S3Build & legacy absorption
S4Integration & interoperability
S5Intelligence, analytics & AI
S6Testing & release mgmt
H1Centre plant & estate
H2Facility hardware custody
H3Connectivity & network
H4Hosting, environments & DR
R1Recruitment & deployment
R2Capacity building & training
R3Governance & committees
R4Adoption assurance & evaluation
How they interlock
Mobilisation weeks 1–13 — leadership on day one
Four interlocks: survey gates everything · specification gates build · training gates deployment · adoption assurance measures all
Five onboarding waves inside months 1–6 — the wave gantt lives in
From sign-off, the Sanjeevani digital backbone deploys to every facility — the domain chips above drill into how each stream delivers
F
The fourteen strategies and their owners
2 marks
Ref.
Strategy
Owner
S1
Requirement gathering & assessment — survey and functional study at every facility
Public Health Specialist with Platform Lead
S2
Specification, architecture & design — signed off before development
Every strategy appears as a stream in the — drill the term to quarters, months and weeks with these streams as the filter.
F
S1–S6 · Software — survey to statewide, in five moves
2 marks
The software path
1Survey & assessment of requirements→2User acceptance test→3Deploy in the pilot facilities — findings incorporated→4UI/UX study→5Deploy for all facilities — in CI/CD mode
🚀From sign-off, the Sanjeevani digital backbone deploys to every facility.
Whatever the Authority signs off after survey & assessment goes to all 11,736 facilities from day one — the pilot, the UAT and the
UI/UX study sharpen it; the CI/CD pipeline carries every release after it.
Each move, briefly
Survey & assessment — requirements captured facility by facility during the assessment months, gating everything downstream
User acceptance test — State users test against the signed-off requirements before anything reaches a citizen
Pilot deployment — deployed in the pilot facilities; findings incorporated before the statewide push
UI/UX study — user-experience research with the functionaries who live in the screens (S2, on the C timeline)
Statewide, in CI/CD mode — every release rides Dev → Staging → Production through the two approval gates
Consolidation closes the six-month window across all 28 districts — exceptions, blackspot closure, final acceptance.
The go / no-go gate before every wave
A facility counts as onboarded only when all five points pass — not on software availability:
Hardware installed and working
Connectivity proven (or offline provision in place)
Personnel trained, device in hand
Master data loaded
First live transactions performed, supervised
Wave by wave — window, load and purpose
Wave
Window
Districts
≈ Facilities
≈ Community workers
Purpose
Pilot
M1 – M2
1
419
2,580
Prove the full chain end to end — mixed urban, rural and tribal terrain; the defect list feeds design before Wave 1 commits
Wave 1
M2 – M3
5
2,096
12,900
First production wave at controlled scale
Wave 2
M3 – M4
8
3,353
20,700
Peak throughput
Wave 3
M4 – M5
8
3,353
20,700
Peak throughput
Wave 4
M5 – M6
6
2,515
15,500
Completion of district coverage
Consolidation
M6
28
residual sites
Exceptions · blackspot closure · final acceptance
Each facility runs a Day −30 → Day +14 sequence inside its wave — connectivity and equipment first, enrolment and
training next, master data and supervised first transactions at the gate, then two weeks of floor support.
G
Risk and Operation & Management Plan
Fourteen material risks owned by name, an operating rhythm that never stops, and service levels met by discipline rather than explained by exception.
Referral-closure & ABHA-linkage measures; the Monday complaints report (received, replied, unanswered — 48-h response on all); implementation progress
Fortnightly
The change board — every candidate change classified before development begins
Monthly
State reviews & coordination committees; SLA report with computation basis and exclusion log; triage-adherence audit; satisfaction sampling; minor release; architecture board
Quarterly
Invoice with progress report; DR drill with restoration validated; preventive maintenance across all sites; digital health campaign; major release; clinical governance forum; security review; steering committee
Definitions first. Measurement definitions are settled in the governance framework before any measurement occurs — definitions settled late are settled adversely.
Evidence at the event. Excluded events (infrastructure outage, third-party failure) are logged, timestamped and notified when they happen — not reconstructed at month-end.
Leading indicators. The floor manages referral closure at day 4 of a 7-day window; the Authority is reported the contracted measure.
Evidence. In one large system only 34.8% of 103,000+ referral attempts ended in a documented completed appointment; closed-loop e-referral lifted documented completion from 79.6% to 95.7%. Referral closure is the measure most likely to be under-estimated by a bidder — and the leak the Centre exists to close.
H
Capacity Building Plan & Quality of Response
Four strategies carry the whole State to productive use — self-learning that personalises itself, leadership webinars with facility hands-on, the training-of-trainers cascade, and citizen advocacy through every medium.
2 marks
Four major strategies — how 105,492 functionaries and 5 crore citizens reach productive use
1Self-learning that personalises itselfSelf-learning modules award a certificate on completion — and then keep working: the modules assess each
functionary's actual usage of the Sanjeevani systems and recommend a personalised training plan for them.
2Webinars by leadership, hands-on at the facilityOnline webinars on Sanjeevani by leaderships and officials, combined with hands-on sessions at the facility
level — the why from the top, the how at the desk where the work happens.
3In-person training through the ToT cascadeTraditional in-person training through training-of-trainers, carried from the State to the facility level —
riding the existing monthly sector meeting so the cascade never needs a new venue.
4Wide advocacy — citizens onto SanjeevaniAdvocacy for citizens to use Sanjeevani through every possible medium — print, social and online media, and
Self-Help Groups street by street.
Full detail
H
The population & the numbers
2 marks
The population, by cadre — 105,492 functionaries
ASHAs41,696
ANMs21,185
PHC / UPHC staff20,302
District staff12,897
CHOs9,412
Centre personnel275
94,943trained every single year
~3×usage lift when trained
4-tiercascade on the sector meeting
48 hresponse on every complaint
QA at three levels — implementation · Centre · adoption
Concurrent evaluation from year three, incorporated
Certification currency and self-learning usage feed the personalised training plans of Strategy 1
H
Who is trained, and how the cascade absorbs it
2 marks
Population
Number
Content spans software · devices · process
Community health activists (ASHA)
41,696
Household capture, high-risk identification, voice agent, offline sync; escalation & counselling; privacy in the household
Platform + STGs, care plans, triage, outbound scripts, disposition discipline
Four tiers: ~30 master trainers → 28 district coordinators + ~112 field trainers → ~2,300 trainers of record at sector level → the workforce.
The monthly PHC-sector meeting already convenes ~63 community workers per sector — one full round is ~3,432 batch sessions ≈ 1.5 sessions per trainer per week. Existing rhythm beats new rhythm.
A trained worker is a productive one — measured by use on the platform, not attendance; ≥90% of frontline workers every year.
H
The evidence base
2 marks
A 28-country survey of 1,141 community health workers found trained workers nearly three times as likely to use digital devices — and three times as likely to believe the tools increase their impact. The authors' conclusion: community health workers are not the obstacle to digital adoption.
Qualitative work in a neighbouring Indian state names six barriers: device and internet access, insufficient training, usability, language and literacy, resistance among senior workers, and no real-time support. Each has a named countermeasure in this plan — shared-device design, the cascade, in-application help, Telugu voice-first interaction, peer trainers of record, and a level-aware assistant with human escalation.
H
Quality of response
2 marks
Concurrent evaluation at the start of year three — an independent look at adoption, incorporated as recommendations, with adoption measured against the pre-evaluation position.
Three QA lanes: implementation quality (per-facility readiness audit), Centre quality (call scoring, triage-adherence, disposition discipline), adoption quality (facility banding with targeted intervention).
Citizen responsiveness: 48-hour response on every complaint; the Monday report names received, replied and unanswered.
Authority capability: analysts operating the governance dashboards independently by year four — capacity building includes the Authority itself.
I
Implementation Schedule
The signed-off solution deploys statewide on day one; the six months then earn every facility the word onboarded — trained, equipped, and with dashboards that match the ground truth.
2 marks
1Survey & assessment — the Authority signs off→2Day 1 — the signed-off solution deploys to all 11,736 facilities→3In parallel — STGs & SOPs customised and configured into Sanjeevani→4Facility onboarded — capacity · hardware · data veracity→5Dashboards for reviews · insights followed up
Statewide on day one
After the survey and assessment period, whatever the Authority has signed off deploys to all facilities on day one — software availability is never the constraint. Meanwhile the customisation and configuration of Standard Treatment Guidelines and Standard Operating Procedures is incorporated into Sanjeevani as a continuous stream.
Drill in ›
What ‘onboarded’ means
A facility is mentioned as onboarded only when capacity building is done, hardware is integrated and placed, and the veracity of ground truth against the data displayed in dashboards is established — the five-point gate in full.
Drill in ›
Dashboards you can govern by
From onboarding, the facility's dashboards are used for reviewing, and actionable insights are reviewed for follow-up — progress displayed State → district → block → facility in real time, not reported.
Drill in ›
I
Waves, districts and throughput
2 marks
Software availability is never the constraint — the platform is deployed to every facility on day one, with STGs
and SOPs configured in continuously. The waves sequence the onboarding work: capacity building, hardware integration and
placement, and the veracity checks that make each facility's dashboards trustworthy for review.
Wave
Districts
≈ Facilities
≈ Community workers
Purpose
Pilot
1
419
2,580
Prove the full chain end to end; the defect list feeds design before Wave 1 commits
Wave 1
5
2,096
12,900
First production wave at controlled scale
Wave 2
8
3,353
20,700
Peak throughput
Wave 3
8
3,353
20,700
Peak throughput
Wave 4
6
2,515
15,500
Completion of district coverage
Consolidation
28
residual sites
Exceptions · blackspot closure · final acceptance
I
The per-facility sequence
2 marks
When
Activity
Monitoring indicator
Day −30 to −21
Facility record confirmed; connectivity provisioned and tested; equipment verified, defects raised
