Contact person: Social Media Screening App for Yoga Course Admissions

Phone:Show

Email:Show

Location: Hyderabad, India

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Title: Build web app to auto-scan public social media handles and generate behavioral screening reports (adaptability vs resistance)

Project summary:
We run online yoga courses and want a tool to help screen applicants for behavioral suitability (e.g. adaptability vs resistance). The system should accept applicant details and public social handles (Instagram, Twitter/X, LinkedIn, TikTok, Facebook public), collect public posts, run automated analyses (sentiment, toxicity, custom classifier for adaptability vs resistance, yoga relevance), and output a PDF/CSV report with scores and flagged posts for human review.

MVP scope (must-have):
Secure admin login and dashboard. Applicant input form with name, email, phone, public social handles, and consent capture. Automated collection of public posts only. NLP pipeline with sentiment analysis, toxicity detection (Perspective API or cloud NLP), custom classifier for adaptability vs resistance. Scoring engine that produces Adaptability score, Resistance risk, Toxicity rate, Yoga relevance percentage. Auto-generated PDF and CSV report. Human review UI to check flagged posts, accept/dismiss, and add notes. Audit log, consent storage, delete-on-request. Admin settings for thresholds and keywords. Basic security: HTTPS, role-based access, encrypted storage.

Nice-to-have (phase 2):
Multi-language support (Hindi, Telugu). Image/video analysis. Batch processing and continuous monitoring. Integration with admission system.

Tech preferences:
Frontend React or Next.js. Backend Python (FastAPI) preferred or Node.js. Postgres database and S3 storage. Use cloud NLP APIs (Google, AWS, Azure) or Hugging Face. Scraping via official APIs where possible. Deployment on AWS/GCP/Azure with Docker.

Deliverables:
Source code in GitHub with README. Deployed staging instance. Admin UI and applicant flow working. Report generation functional. Test dataset with 10 sample applicants processed. Short admin manual.

Acceptance criteria:
System extracts at least 90 percent of public posts for provided handles. PDF report includes header, scores, and flagged posts with links. Consent record stored and retrievable. Data deletion works within 7 days. HTTPS and no plaintext secrets. Demo video of workflow.

Budget and timeline:
Typical MVP budget USD 3000–6000 (approx INR 2.5L– 5L). Delivery preferred in 4–8 weeks.

Proposal requirements:
Short technical approach. Portfolio samples. Estimate and timeline. Preferred stack and API costs. Willingness to sign NDA and provide support. Answer screening questions: 1) How will you ensure only public data and respect TOS? 2) Experience with Perspective API or NLP? 3) Handling of Hindi/Telugu posts?

Evaluation:
Shortlisted candidates will be given a small paid test. The test is to build a script that accepts 3 Twitter handles, fetches last 100 public tweets, runs sentiment and toxicity analysis, and generates a PDF report. Payment USD 100, deadline 5 days.

Legal and ethical:
Only public posts may be collected. Applicant consent must be captured. AI output is advisory only, final decisions involve human review. Privacy and fairness must be built in." (client-provided description)