Python Tender Aggregator SaaS Build
₹12500-37500 INR
Posted 1 minute ago
•
Ends in 6 days
₹12500-37500 INR
Paid on delivery
I am developing a Software-as-a-Service platform that automatically gathers government tenders from multiple countries, with India as the first live market. The stack is Python, Django and PostgreSQL, and the end product must offer a smooth sign-up/login flow, a paid subscription model and concise AI-generated summaries for every tender the system captures. Core workflow • A crawler schedules itself, visits the main Indian government tender portals, parses new notices in near-real time and stores them. • An AI micro-service (OpenAI, Hugging Face or similar) distills each notice into a short, business-friendly summary while tagging industry, location and deadline. • Authenticated users browse or search tenders through a responsive dashboard and receive daily email digests filtered by their saved preferences. • The payment module (Stripe, Razorpay or another mainstream gateway) protects premium endpoints and renews subscriptions automatically. Deliverables 1. Git-based Django project ready for Docker deployment, including reusable scraping jobs, Celery (or APScheduler) tasks and unit tests. 2. Admin and user dashboards with role-based permissions, JWT or session authentication, password reset and social login optionality. 3. Fully integrated payment workflow: plan creation, checkout, webhook handling and subscription tier enforcement. 4. AI summarisation service with a fallback to manual review. 5. Clear setup guide and minimal DevOps script (Docker Compose or Ansible) so I can spin up staging and production quickly. Acceptance criteria • Scraper captures 90 %+ of new Indian government tenders within one hour of publication. • Summaries stay under 120 words, keep all critical data and pass a random manual audit. • Subscription enforcement blocks premium routes for unpaid users. • Codebase follows PEP 8, is covered by at least 70 % automated tests and ships with README instructions that let me reproduce the deployment on a fresh VPS in under two hours. When you apply, please attach a detailed project proposal explaining your architecture choices, timeline and previous experience with similar data-heavy SaaS builds. I will shortlist based on the clarity of that proposal and proof you have tackled large-scale scraping or AI summarisation before.
Project ID: 40830425
About the project
11 proposals
Open for bidding
Remote project
Active 5 secs ago
Place your bid
Bid amount Email addressBenefits of bidding on Freelancer
Set your budget and timeframe
It's free to sign up and bid on jobs
11 freelancers are bidding on average ₹22,273 INR for this job


Hi! I'm Fernando from Mexico. The real challenge with a tender aggregator isn't simply collecting notices, but keeping the data accurate, avoiding duplicate records, and ensuring users receive relevant opportunities before deadlines pass. I'd build this with Django and PostgreSQL, using Celery for scheduled scraping and background processing. Each portal would have an independent scraper so new countries and sources can be added without changing the core system. I'd also implement AI summaries with structured industry, location and deadline extraction, plus manual review when confidence is low. The platform would include user dashboards, saved search preferences, automated email digests, and Razorpay or Stripe subscription handling with webhook-based access control. Docker Compose, automated tests and deployment documentation would be included. I'd suggest starting with a reliable India-focused MVP, then expanding portal coverage based on measured scraping accuracy. Best regards, Fernando


The main challenge is making tender capture dependable across portals with different page structures, while keeping each notice traceable from scraping through summarisation, search, and subscription access. A single generic crawler is likely to miss changes or create duplicates, so the India launch should use portal-specific adapters and a shared processing pipeline. I’d structure the Django app around tender records, users, saved filters, and subscription entitlements. Scheduled Celery jobs can fetch and parse notices, deduplicate them, retain source links and capture timestamps, and retry failures without blocking other portals. Summarisation should run as a separate service step, enforce the 120-word limit, and store industry, location, and deadline alongside the generated text; uncertain or failed results can be flagged for manual review. I’d also make capture and summary quality measurable against sampled notices rather than assuming the stated accuracy targets are met from day one. For the user workflow, I’d keep authentication, saved preferences, tender search, and premium access rules explicit in Django. Payment checkout and webhook handling should be idempotent so repeated gateway events cannot incorrectly change a subscription. Daily digests can use the same saved-filter logic as dashboard search. Responsive templates, clear empty/error states, focused unit tests for parsing and entitlement rules, and Docker Compose setup would keep the first release practical to operate and extend. I’d keep the first implementation focused on the India workflow, with reusable scraper and task boundaries so additional markets do not require rewriting the core. Git deliverables would include setup and deployment instructions, tests for the critical paths, and a manual-review route for AI output that needs checking. Which tender portals are in the initial scope, and can you provide representative notice pages or sample data for their formats?


HELLO, "Government tender aggregator" - you need reliable multi-portal scraping, AI summaries and paid subscriptions working together in one SaaS platform. I'll build the Django and PostgreSQL backend around reusable portal-specific scrapers, Celery scheduling and structured tender storage, then connect AI summarisation, searchable dashboards, user authentication and subscription payments. I'll include webhook-based payment verification, Docker deployment, automated tests and setup documentation. One important detail is handling inconsistent portal formats, duplicate notices and changing publication dates so the crawler doesn't miss tenders or send repeated alerts. I'll handle the complete implementation and keep the architecture ready for additional countries. Which Indian government tender portals should be included in the first release? I'll provide a clear architecture and implementation plan with the deliverables and testing approach. Thank you for your consideration. I look forward to discussing the build. THANKS


Hi, The main challenge is reliably collecting tenders from multiple government portals, processing them quickly, and turning unstructured notices into searchable, useful business intelligence. I'd build this with Django, PostgreSQL, Celery, Redis, and Docker. Each government portal would have an independent scraping adapter, allowing sources to be maintained without affecting the rest of the system. Scheduled workers would detect new notices, normalize records, remove duplicates, and track publication timestamps. For AI summaries, I'd use OpenAI with structured outputs to extract deadlines, industries, locations, eligibility, and key requirements into consistent summaries under 120 words. Failed or uncertain extractions would enter a manual review queue. I'd also implement subscription management with Razorpay/Stripe, webhook-based payment verification, role-based access, personalized email digests, and automated tests. Would you like the initial release to cover only central Indian government portals or state-level portals as well?


Hi You’re building a Django SaaS that captures Indian government tenders quickly, summarizes them, and gates premium access by subscription. I’d first map portal formats and publication signals, then build reusable Python scrapers with Celery, PostgreSQL deduplication, and monitored retries. For summaries, I’d keep critical fields separate from generated text and route uncertain results to manual review. That’s where tender data can get messy. Our Django news aggregator ITK used ML classification and real-time alerts, relevant to tender tagging and timely updates. For the rollout, I’d sequence ingestion and tests first, then summaries, subscriptions, and dashboards. Kindly help me pin down two details: 1. Which Indian tender portals are essential for the first release, and do you have sample notices or portal access details? 2. Should daily digests use a fixed send time, or each user’s local time zone? Happy to share a quick plan in chat. Regards, Gayatri
内容来自 Freelancer News,版权归原作者所有。
阅读原文 ↗