Aastha Tiwariprovide me a website
Sign in to leave a comment
PROJECT: GENDERISTI AI
AI-Powered Government Fund Transparency & Ground Reality Monitoring Platform
Tagline
"Every Citizen's Voice Matters. Every Public Fund Deserves Accountability."
Build a complete, modern, responsive, AI-powered web application called Genderisti AI.
Example:
Project: Village Road Development
Sanctioned Budget: ₹10,00,000
Reported Expenditure: ₹8,50,000
Remaining Amount: ₹1,50,000
Reported Progress: 100%
FINAL MISSION
Build Genderisti AI as a citizen-centric public development transparency platform.
The platform should not simply show how much money was sanctioned or spent. It should help people understand whether that money resulted in actual development.
Final tagline: "From Public Funds to Public Progress — Every Rupee, Every Project, Every Voice Matters."
This platform aims to improve transparency and accountability in public development projects by connecting government fund allocation, expenditure records, physical development progress, AI-powered object detection, and citizen participation.
The platform should help citizens understand:
How much government funding was sanctioned for a development project?
Which department or administrative authority received the funds?
Where was the money allocated?
Which contractor or agency was assigned the work?
How much money was actually spent?
The unique innovation of Genderisti AI is the combination of three intelligence layers:
Layer 1: Financial Intelligence
Tracks sanctioned funds, allocation, release, and expenditure.
Layer 2: Visual Intelligence
Uses AI object detection and image comparison to analyse visible development work.
Layer 3: Citizen Intelligence
Collects public feedback, photographs, videos, and complaints.
All three layers must connect to the same development project.
Was the promised development work completed?
Does the physical condition match the official project report?
Can citizens report incomplete or poor-quality work?
The platform must connect financial information with real-world visual evidence.
Create a complete government fund management module.
Store the following information:
Scheme or project name
Government department
State, district, block, and village
Sanctioned amount
Fund release date
Fund allocation details
Implementing authority
Responsible officer
Contractor or agency
Approved project budget
Actual expenditure
Remaining amount
Payment history
Supporting documents
Create a financial flow:
Government Fund Sanctioned → Department Allocation → Administrative Approval → Contractor Assignment → Fund Release → Development Work → Evidence Upload → AI Verification → Citizen Feedback.
Display sanctioned, allocated, released, and spent amounts separately.
Support these public development categories:
Infrastructure and roads
Education and schools
Healthcare facilities
Cleanliness and sanitation
Drinking water supply
Environmental protection and greenery
Public spaces and parks
Social welfare
Sports facilities
Each project must have its own budget, location, timeline, responsible authority, contractor, and evidence records.
The system should analyse uploaded photographs and videos of development sites.
Detect:
Road Infrastructure
Potholes
Road cracks
Damaged footpaths
Broken streetlights
Damaged public infrastructure
Cleanliness
Garbage piles
Overflowing dustbins
Scattered waste
Show detected objects with bounding boxes, labels, confidence scores, and image annotations.
Allow users to upload images and videos from their devices.
Create an AI-based development verification module.
For every project, store:
Before-work photographs
During-work photographs
After-work photographs
Project location
Project dates
Approved budget
Reported expenditure
Contractor details
Project: Village Road Construction
Sanctioned Amount: ₹10,00,000
Reported Work: Road construction completed.
Compare before-and-after photographs to identify visible improvements.
Example:
AI Analysis:
Road surface detected.
Visible cracks identified.
Possible damaged sections highlighted.
Before-and-after comparison generated.
Display the result as:
Visible improvement detected
Partial improvement detected
No clear improvement detected
Create a unique feature called Fund-to-Ground Intelligence.
Connect each sanctioned fund with its actual development project and visual evidence.
AI Visual Analysis:
Road construction visible.
Some cracks detected.
Possible incomplete sections identified.
Citizen Feedback:
Show financial records, project progress, AI analysis, and citizen feedback together.
Citizen Feedback:
System Status:
System Status:
Allow citizens to:
Allow citizens to:
Allow citizens to:
Allow citizens to:
Allow citizens to:
Allow citizens to:
Allow citizens to:
Report misuse-related concerns.
Submit complaints with location details.
Track complaint status.
View official responses where available.
Each complaint must receive a unique Complaint ID.
Complaint statuses:
Submitted
Under review
Assigned
Investigation in progress
Response received
Resolved
Rejected with explanation
Escalation recommended
Build an AI-powered evidence analysis system.
Features:
Duplicate image detection
Similarity checking
Image metadata inspection
Possible image manipulation indicators
Project-image relevance checking
Missing evidence detection
Before-and-after image comparison
Object detection
Generate warning flags for suspicious or inconsistent evidence.
AI results must be treated as preliminary indicators, not conclusive proof of fraud or wrongdoing.
Create separate dashboards for administrative authorities and contractors.
Display:
Assigned projects
Approved budgets
Funds released
Reported expenses
Work completion status
Uploaded invoices
Project photographs
Pending work
Citizen complaints
Record who approved, updated, or verified each project.
Where authorized financial integrations exist, support payment verification using official transaction references.
Do not treat screenshots alone as independently verified payments.
Create a dashboard showing:
Total sanctioned funds
Total funds released
Total expenditure reported
Remaining funds
Active projects
Completed projects
Delayed projects
Projects awaiting verification
Citizen complaints
Resolved complaints
Projects with missing evidence
Projects flagged for manual inspection
Include charts, maps, category-wise spending, and district-wise project information.
Give citizens a voice through:
Public project feedback
Development priority suggestions
Community polls
Complaint submissions
Project ratings
Public issue reporting
Where applicable, allow citizens to suggest development priorities for their village or locality.
Generate downloadable reports:
Project-wise fund report
Department-wise expenditure
District-wise development report
Contractor-wise project report
Delayed project report
Citizen complaint report
AI evidence analysis report
Support PDF and Excel exports.
Maintain an audit trail for financial and project updates.
Implement role-based access for:
Citizen
Government Department Administrator
Authorized Government Officer
Contractor
Platform Super Administrator
Each role should have appropriate access permissions.
Integrate an interactive map.
Display:
Project locations
Village and district boundaries where available
Development project status
Reported infrastructure issues
Citizen complaints
Project progress
Allow users to filter projects by location, category, department, and status.
Design requirements:
Main navigation:
Dashboard
Government Funds
Departments
Development Projects
AI Object Detection
Before-and-After Verification
Citizen Complaints
Start by creating the complete project structure, database schema, frontend, backend, and working dashboards. Then implement the modules step by step, test all important workflows, and provide clear setup instructions.
Mercy-Provide serves five human roles, each scoped by role-based access permissions; detailed rules live in Functional Requirements.
| Role | Responsibilities and goals | Inputs | Decisions / observable success | Access |
|---|---|---|---|---|
| Citizen | Understand how much funding was sanctioned, which department received it, where it was allocated, which contractor or agency was assigned, and how much was actually spent; judge whether promised work was completed and whether physical condition matches official reports; raise concerns and shape local priorities | Project searches, photographs and videos, location details, complaints, feedback, ratings, priority suggestions, poll votes | Can trace a project from sanction to spend to visible outcome; complaint status and official responses are trackable | Register/log in; search and view public project information; upload media; submit and track complaints; give feedback, ratings, priority suggestions; vote in community polls |
| Government Department Administrator | Oversee departmental projects and fund flows | Assigned projects, approved budgets, funds released, reported expenses | Departmental dashboard reflects assigned projects, approved budgets, funds released and reported expenses | Department dashboard |
| Authorized Government Officer | Verify reported progress against physical evidence and respond to citizen reports | Assigned projects, budget and expense figures, uploaded visual evidence, complaints | Discrepancies between official reports and physical condition are identified; official responses issued | Officer dashboard covering assigned projects, approved budgets, funds released, reported expenses |
| Contractor | Execute assigned work and account for expenditure | Assigned projects, approved budgets, funds released, reported expenses | Project records show assigned work and reported expenses | Contractor dashboard |
| Platform Super Administrator | Govern platform-wide access and role permissions | Role assignments, permission settings | Each role holds appropriate access permissions | Platform-wide administration |
Primary journey (Citizen). Register and log in → search a development project → view public project information (sanctioned amount, receiving department, allocation location, assigned contractor or agency, actual spend) → compare with physical evidence → upload photographs/videos and submit a complaint with location details → track complaint status and view official responses → contribute feedback, ratings, priority suggestions and poll votes.
Supporting actors. AI-powered object detection and the platform itself support verification and transparency; they are not human personas.
Default: persona names beyond the five listed roles, and any additional duties or permissions, are not established by current evidence.
Each journey below is independent; none requires completing another first. All three intelligence layers (financial, visual, citizen) attach to the same project record.
Actor: Citizen (public view); Officer/Contractor (their own records). Trigger: Citizen wants to know what was sanctioned and whether it was spent.
Result: Citizen can judge whether sanctioned money maps to a real project and visible work. Failure/recovery: If evidence is missing, the project is flagged as missing evidence and surfaced for manual inspection. Continuation: Citizen proceeds to visual verification (Journey 2) or files a complaint (Journey 3).
Actor: Citizen or Officer. Trigger: A project claims completion or progress that the user wants to check.
Result: A preliminary visual assessment (e.g., road construction visible, some cracks detected, possible incomplete sections) shown alongside financial records and citizen feedback. Constraint: AI results are preliminary indicators, not conclusive proof of fraud. Failure/recovery: Flagged or low-confidence evidence routes the project to manual inspection. Continuation: User may attach findings to a complaint.
Actor: Registered Citizen. Trigger: Resident observes incomplete work, poor-quality construction, or misuse-related concerns.
Result: Complaint is linked to the project's financial and visual record; resolved complaints count toward the government dashboard. Failure/recovery: If the complaint cannot be matched to a project, it remains a public issue report pending location-based review. Continuation: Citizen may also rate the project or submit feedback.
Actor: Citizen. Trigger: Citizen wants a say in local spending.
Result: Feedback aggregates into the project's citizen-intelligence layer. Continuation: Suggestions inform future prioritization; no approval step is implied.
Actor: Administrative authority or Contractor. Trigger: Routine review of assigned work.
Result: Auditable trail of actions per project. Continuation: Updated evidence re-enters AI verification (Journey 2).
Actor: Government administrator. Trigger: Periodic oversight.
Result: Consolidated view across all three layers. Continuation: Flagged projects feed manual inspection.
Greenfield Village road project (Rural Development): ₹10,00,000 sanctioned, ₹8,50,000 reported expenditure, ₹1,50,000 remaining, status "completion claimed, verification pending" — used to exercise Journeys 1–3 end to end.
[Design proposal; unspecified choices are defaults.]
Visuals, Colours and Theme
mercy-provide is a civic-tech transparency platform for two audiences at once: citizens checking whether a sanctioned road, drain or streetlight actually exists, and district officials and contractors answering for it. The register is trust without warmth-theatre — legibility of money, place and evidence. The muse is Erik Spiekermann: typography as public infrastructure, signal colours used like transit lines, numbered categories, pictograms always paired with type. The brief's "premium government-tech dashboard" is honoured through order, ruled columns and real photographic evidence, not generic dashboard chrome.
Palette (light mode, AA at body size): background #F4F1EA (warm paper), surface #FFFDF8 (off-white panels), text #1A1A18 (near-black ink), primary #1F4E3D (deep transit green — navigation, headers, primary action), accent #D93A1E (signal red-orange — reserved for discrepancies, pending verification and complaint flags, never decoration), muted #6E6A60. Category line-codes: mustard #C9922B (waste), slate blue #2C5F7C (water), ochre #8A6D3B (environment), green #1F4E3D (infrastructure). Proportions: 70% paper/ink, 20% green structure, 10% signal colour. Hairline rules #DAD4C7.
Typography: Fira Sans throughout — headings SemiBold/Bold, tracking -0.02em, sentence case, flush-left ragged-right, never centred; small-caps Fira Sans for section labels and table headers (letter-spacing 0.08em, 12–13px). Numbers in tabular figures; rupee figures carry the largest weight in any row. Scale 1.25 modular with display jumps: 20 / 22 / 28 / 36 / clamp(44px, 8vw, 112px) hero; body 17px/1.6, table 15px, micro-label 12px small caps. No Inter, Roboto, Poppins or system-ui.
Shape language: rectilinear and honest — 2px radius on inputs and buttons, 0px on cards and tables, 1px hairline rules as the main structural device, 6px colour-coded left borders marking project category and status. Pictograms on a 24px grid, 2px stroke, always with text labels. No soft shadows beyond a single 1px offset line; status chips are rectangles with 2px radius.
Imagery: real, current, geotagged site photography is the evidence layer — before/after photos of roads, drains, streetlights, waste bins and water points, full-bleed inside ruled frames with a caption bar carrying date, district and AI detection confidence. AI detection panels overlay thin signal-red bounding boxes with small-caps labels and confidence scores. Maps are schematic and line-coded by department colour, not satellite decoration. No stock people, no illustration, no gradients. Sample demonstration data is clearly labelled fictional.
Layout: 12-column ruled grid with a persistent 240px left sidebar carrying numbered navigation (01 Overview, 02 Funds, 03 Projects, 04 Visual Evidence, 05 Complaints, 06 Map) and a top strip with district selector and role. Content is tabular: fund rows aligned label-left / value-right on a shared baseline, horizontal rules between every row. Mobile collapses to one column with the sidebar as a numbered drawer; tables become stacked label/value pairs, never horizontally clipped. Headlines, labels, numbers and controls stay whole at 375px, 768px and 1280px.
Create a premium, professional government-tech dashboard.
Design requirements:
Clean modern interface
Responsive layout
Sidebar navigation
Professional typography
Design requirements:
Interactive charts
Financial summary cards
Project tables
Image galleries
AI detection panels
Map visualisation
Complaint tracking timeline
[Design proposal; unspecified choices are defaults.]
The public entry is a ruled editorial hero, not a dashboard card. The left 8 columns hold an oversized flush-left Fira Sans statement — "₹10,00,000 sanctioned. Is there a road?" — at clamp(44px, 8vw, 112px), with a small-caps line-code label above it: "PROJECT 04 · VILLAGE ROAD DEVELOPMENT · DISTRICT PUNE". Directly beneath, a full-width ruled fund-to-ground band runs edge to edge with four aligned tabular figures — sanctioned, released, spent, verified on ground — each with a tiny pictogram and a green or signal-red verification mark, separated by 1px vertical rules and reading as a transit timetable for money. The right 4 columns are a single real site photograph in a hard-edged frame with a caption bar and one AI detection box. Paper ground, ink type, green structure, one red flag. No centred headline, no gradient, no floating cards.
Signature moves:
These moves carry the three connected intelligence layers — Financial, Visual and Citizen — onto the same project record, and surface discrepancies as "flagged for human review", never as corruption.
[Design proposal; unspecified choices are defaults.]
Interaction Model: Static (direction) Motion Tempo: restrained Hero Dimensionality: flat
Landing Hero Motion Brief
Focal subject: the real site photograph in its hard-edged frame, paired with the ruled fund-to-ground band beneath the statement. Input → transformation → outcome: as the band scrolls into view, the four rupee figures count up once from zero to their sanctioned, released, spent and verified values, and each verification mark resolves from neutral to green or signal red — the citizen's outcome is a legible, settled ledger, not an animation.
Motion vocabulary: 120–180ms ease-out on drawer open, tab switch and status chip change; a single count-up per financial strip on first scroll into view; no parallax, no decorative loops, no hover-lift. The hero's first frame is fully composed and static — statement, line-code label, four figures and photograph all present before any motion.
Reduced-motion state: prefers-reduced-motion removes counting and transitions entirely, leaving instant final values and states; the hero remains complete and readable. Scrollable rows wrap or scroll so every item can be brought fully into view.
Private personal information must not be exposed publicly.
Implement:
Secure authentication
Role-based access control
Database integration
Backend APIs
Image and video storage
Secure document upload
Audit logs
Data validation
Error handling
Backup support
Privacy protection
Integrate a Computer Vision system using YOLO or another suitable object detection model.
Use a suitable technology stack supported by the development environment.
INSTRUCTIONS FOR GOOGLE ANTIGRAVITY IDE
Act as a senior full-stack developer, AI/ML engineer, UI/UX designer, and software architect.
Do not build only a static landing page or UI prototype. Develop a functional application with a proper frontend, backend, database structure, authentication, dashboards, and AI integration architecture.
Evidence insufficient
Manual inspection required
AI must not automatically declare corruption or fraud. It should identify possible discrepancies for human investigation.
AI results must be treated as preliminary indicators, not conclusive proof of fraud or wrongdoing.
Create a fictional demonstration project:
Project: Village Road Development
Location: Greenfield Village
Department: Rural Development
Sanctioned Budget: ₹10,00,000
Reported Expenditure: ₹8,50,000
Remaining Balance: ₹1,50,000
Project Status: Completion claimed, verification pending.
Generate realistic sample project data, charts, complaints, and AI analysis examples.
Clearly label all fictional data as DEMO DATA.
Never present simulated payments, government approvals, or AI analysis as genuine official records.
If an actual AI model, government API, or banking integration is not available, create a clearly labelled integration-ready module rather than pretending the service is connected.
Water Supply
Visible water leakage
Waterlogging
Damaged water infrastructure
Overflowing water tanks where visually identifiable
Environment
Trees
Plants
Green areas
Visible vegetation changes
Damaged or dry plants
Public Infrastructure
Benches
Playground equipment
Public buildings
Sports infrastructure
Visible construction or maintenance defects
Contractor Management
Performance Analytics
Reports
User Management
Settings
Every citizen’s voice matters. Every public fund deserves accountability.
Village Road Development
Fund-to-ground · Three intelligence layers
Demo dataTracks the money, rupee by rupee.
Financial Intelligence follows sanctioned funds, allocation, release and expenditure for each development project, so the public can read what was promised against what was actually spent.
Checks the work on the ground.
Visual Intelligence uses AI object detection and image comparison to analyse visible development work in site photographs, including before-and-after views of the same location.
Collects what people see.
Citizen Intelligence collects public feedback, photographs, videos and complaints about incomplete work, construction quality or misuse concerns, reported against the location and project.
All three layers connect to the same development project. Financial, visual and citizen evidence are read together.
Bound to one project recordmercy-provide should not simply show how much money was sanctioned or spent. It should help people understand whether that money resulted in actual development — reading the financial record, the visual evidence and the citizen’s account against the same project, without spin.
From Public Funds to Public Progress Every Rupee, Every Project, Every Voice Matters.
AI caution
AI analysis results are preliminary indicators, not conclusive proof of fraud or wrongdoing. Discrepancies are surfaced for human investigation; no automatic accusation is made.
Every figure, project, complaint and AI analysis shown on this page is fictional demonstration data, labelled DEMO DATA. None of it is presented as a genuine official record, and it should not be read as a statement about any real department, contractor or citizen.
AI models, government APIs and banking integrations that are not available in this build appear as clearly labelled integration-ready modules. They are shown as planned interfaces, not as connected services processing live data.
Every citizen’s voice matters. Every public fund deserves accountability.
Village Road Development
Fund-to-ground · Three intelligence layers
Demo dataTracks the money, rupee by rupee.
Financial Intelligence follows sanctioned funds, allocation, release and expenditure for each development project, so the public can read what was promised against what was actually spent.
Checks the work on the ground.
Visual Intelligence uses AI object detection and image comparison to analyse visible development work in site photographs, including before-and-after views of the same location.
Collects what people see.
Citizen Intelligence collects public feedback, photographs, videos and complaints about incomplete work, construction quality or misuse concerns, reported against the location and project.
All three layers connect to the same development project. Financial, visual and citizen evidence are read together.
Bound to one project recordmercy-provide should not simply show how much money was sanctioned or spent. It should help people understand whether that money resulted in actual development — reading the financial record, the visual evidence and the citizen’s account against the same project, without spin.
From Public Funds to Public Progress Every Rupee, Every Project, Every Voice Matters.
AI caution
AI analysis results are preliminary indicators, not conclusive proof of fraud or wrongdoing. Discrepancies are surfaced for human investigation; no automatic accusation is made.
Every figure, project, complaint and AI analysis shown on this page is fictional demonstration data, labelled DEMO DATA. None of it is presented as a genuine official record, and it should not be read as a statement about any real department, contractor or citizen.
AI models, government APIs and banking integrations that are not available in this build appear as clearly labelled integration-ready modules. They are shown as planned interfaces, not as connected services processing live data.
No comments yet. Be the first!