Panipat City (Haryana) • AC-25
- Live Booths
- 204
- Margin Threat
- High
Platform · mandata.ai
mandata.ai is the constituency management platform from Barnstorm Co-operations: Jan-Sunwai workflow management, citizen connect, booth intelligence, and MPLADS / MLACDS tracking — under role-based access, activity history, and human oversight.
What it is
mandata.ai is not a marketing cloud and not an unsupervised bot. It is a political operations platform for India: households, booths, cases, outreach, and sanctioned works under permissioned access. Campaigns use it as the booth file that argues back. Officeholders use it as Jan-Sunwai workflow management, citizen connect, and fund discipline.
18.2M
Citizens on licensed instances
Households and electors currently on licensed mandata.ai instances in the sample operating set — not a national claim.
19 hrs
Median Jan-Sunwai first response
Median time from grievance intake to first logged human action on live instances in the sample set.
Same night
Field write-back cadence
Target operating cadence: contacts and notes reconcile to the constituency graph before the next principal briefing.
41
Live constituency instances
Assembly and parliamentary units currently on a licensed mandata.ai instance in the sample operating set.
Madhya Pradesh
2018, 2023 Assembly | 2019, 2024 Lok Sabha
6 Lok Sabha seats · 30 Assembly segments
Teal states are live theatres · tap for the cycle file
Who it is for
mandata.ai is built for sitting MPs, sitting MLAs, constituency offices, party operations, campaign command, and large public-facing political offices.
MP constituency management: Jan-Sunwai, citizen connect, MPLADS tracking, and booth intelligence on one graph.
MLA constituency management at assembly grain — MLACDS tracking, grievance workflows, presence against the map.
Office-level operational discipline: named roles, SLAs, and a file the desk can close against.
A political operations platform for India: common graph, named access, audit trail — under the party’s own bench.
Booth file, contacts, and IDs that write back the same night so political campaign analytics are usable before noon.
High-volume grievance and outreach environments that cannot run on a shared password and a WhatsApp group.
For campaigns
The booth file has to argue back. Booth universes, contacts, and IDs write to the same graph as Jan-Sunwai and paid. At 7 a.m. the file names which booths moved — and what to do before noon.
For officeholders
Governing is a graph. Jan-Sunwai, MPLADS / MLACDS, and citizen connect sit on the household. The next race is already in the inbox. mandata.ai makes neglect visible early enough to be expensive to ignore.
For parties
A common graph. Named access. Party operations teams license mandata.ai as infrastructure — booth intelligence, outreach, audit trails — under their own bench, or sit with Barnstorm Co-operations advisory where the race is live.
Modules
01
Booths, households, and institutions as a living constituency intelligence layer — not a CSV that went stale on Friday.
Who uses it
Campaign desks, constituency offices, party operations teams.
What visibility it improves
A single operating picture: who lives where, which booth, which grievance, which sanctioned work.
02
Grievance redressal system for sitting MPs and MLAs: intake, routing, SLA, write-back. The person at the door is already in the file.
Who uses it
Officeholders and grievance desks running Jan-Sunwai workflow management.
What visibility it improves
First-response time, named assignment, and whether neglect is still invisible on Monday.
03
A citizen connect platform against the graph: permissioned outreach for presence, scheme camps, and household contact — not a blast list.
Who uses it
Communications and field, under role-based send rights.
What visibility it improves
Outreach that can be audited, scoped to a universe, and revoked.
04
MPLADS tracking for MPs and MLACDS tracking for MLAs, tied to work and ward: what is sanctioned, what is stuck, what is visible on the ground.
Who uses it
Sitting MPs (MPLADS) and MLAs (MLACDS), with office staff on named access.
What visibility it improves
Whether a sanctioned work is still a rumour in the constituency.
05
Booth-level intelligence: net favorability, turnout, and weekly contacts at booth grain, with nightly field write-back.
Who uses it
Campaign command and sitting offices that still think in booths.
What visibility it improves
Which booths moved — and what to do before noon.
06
A one-page note for principals: booths that moved, Jan-Sunwai that breached SLA, works that stalled.
Who uses it
Principals and chiefs of staff.
What visibility it improves
Decision support without a forty-slide deck.
Operating workflows
Jan-Sunwai lands on a household, then a booth, then a named desk. SLA is a clock. The grievance redressal system makes delay visible before it becomes a political fact.
Which booths moved, which cases breached, which works are stuck — one operating picture for the principal, not five vendor dashboards.
Citizen connect against the graph, permissioned by role. Presence and scheme camps as planned work, not blasts from a shared phone.
Contacts, IDs, and notes write the same night. A briefing that cannot be contradicted by a booth agent is folklore.
MPLADS tracking and MLACDS tracking: sanctioned, stalled, visible on the ground — against ward and work, not a spreadsheet in a drawer.
The race file is the office file. Counting day is a handoff. The household in the campaign is the household in the inbox.
Operating controls
mandata.ai holds political and constituency data that can damage a principal if it leaks, drifts, or is used without authority. Barnstorm Co-operations treats access as an operating problem: who may see, who may send, who may close a case, and what the log will show on Monday. This is office-level discipline, not a legal appendix.
Principal, chief of staff, booth in-charge, grievance desk — named roles, not a shared password. Sensitive races are compartmented. A constituency management platform that anyone can open is not a platform.
Citizen connect is not a blast list. Send rights are granted, scoped to a universe, and revocable. Unpermissioned outreach does not leave the console.
Case notes, assignments, outreach, and fund-status changes write an audit trail. Accountability is a log a chief of staff can read, not a meeting that never happened.
Assistance may rank, route, and draft briefings. It does not publish, sanction, or close a case without a named operator. Models are staff, not oracles.
Jan-Sunwai SLAs, presence calendar, and MPLADS / MLACDS status are visible to the roles that own them. Neglect is expensive because it is visible.
Constituency operations run as a controlled environment: who is in the file, who touched it, and whether the booth sheet still matches the graph.
Live races and sensitive cases stay inside the room that owns them. Barnstorm Co-operations will not sit both sides of a contest. Platform licenses in a contested constituency are disclosed before contract.
How it is bought
For parties, large offices, and firms with a consulting bench already. mandata.ai underneath your operators — constituency graph, workflows, access control. Choose this when you already have a room and need the graph.
For first-time principals and rebuilds. Barnstorm Co-operations sits the room and stands up the instance. Combined engagement is the default when the file does not yet exist. Choose this when campaign-to-governance continuity is the brief.
Advisory practiceQuestions principals actually ask
Sitting MPs and MLAs, constituency offices, party operations teams, campaign command structures, and large public-facing political offices that need a constituency operating system — not a blast list.
Yes. Parties, large offices, and firms with a bench can license mandata.ai alone. Everyone else usually wants the Barnstorm Co-operations room as well.
Intake, routing, SLA, and write-back onto the household and booth. Median first response is a managed metric, not a slogan. Human operators close cases; the system makes neglect visible.
Sanctioned works are tracked to ward and file: what is approved, what is stuck, what is visible on the ground. Sitting MPs use MPLADS tracking; sitting MLAs use MLACDS tracking — both under named access.
It is a constituency management platform with CRM-like outreach — permissioned, logged, and scoped by role. It is a political operations platform for India, not a generic marketing cloud.
Routing and briefing inside a controlled console. Models do not publish, sanction, or close a case without a named operator. Human oversight is an operating rule, not a disclaimer.
Access
mandata.ai for MP and MLA constituency management, with governance, access, and accountability a serious office can sit with.