Cloud-Native AI Platform for Automated Tender Intelligence, Compliance Generation & Process Auditing (DGCP – Dominican Republic)
Budget: $750 – $1,500 USD
I’m looking for a seasoned developer (or small team) to turn the idea of a cloud-native AI platform into a working product that automates tender intelligence for the Dominican Republic’s DGCP portal, with an initial focus on Technology-sector bids.
What I need the platform to do
• Automatically ingest DGCP tender notices, specifications, addenda, and Q&A.
• Use NLP to analyze each document, extract key dates, requirements, evaluation criteria, and risk flags, then present the insights in a clean web-based dashboard (our only required output format).
• Generate pre-filled compliance documentation that aligns our internal policies with each tender’s mandatory clauses.
• Run lightweight process audits—flagging missing approvals, outdated certifications, or gaps between the tender’s stated rules and our compiled response file.
Core expectations
• Cloud-native stack: microservices, Docker containers, CI/CD pipeline, and infrastructure-as-code so we can deploy painlessly on AWS or GCP.
• Modern language tooling is up to you—Python, Node.js, or Go—provided the NLP pipeline remains transparent and easy to extend.
• Model layer should leverage reputable LLMs (OpenAI, Hugging Face, or similar) while allowing us to swap providers if policy or cost changes.
• Dashboard must support full-text search, filter by deadline, one-click export of compliance packs, and an audit trail of every action.
• Security: role-based access, encrypted data at rest, and logging that meets ISO 27001 style traceability.
Deliverables
1. Source code in a private Git repository with clear README and setup scripts.
2. Container images and IaC templates.
3. Initial trained or fine-tuned NLP models.
4. Working demo with at least three recent DGCP Technology tenders, showing the full flow from ingestion to compliance pack and audit report.
5. Brief hand-off session plus 30-day post-delivery bug fix window.
If you have built tender, RFP, or contract-analysis tools before, that’s a big plus. Please outline your proposed tech stack, any similar projects, and the timeline you’d need for a first functional release.
What I need the platform to do
• Automatically ingest DGCP tender notices, specifications, addenda, and Q&A.
• Use NLP to analyze each document, extract key dates, requirements, evaluation criteria, and risk flags, then present the insights in a clean web-based dashboard (our only required output format).
• Generate pre-filled compliance documentation that aligns our internal policies with each tender’s mandatory clauses.
• Run lightweight process audits—flagging missing approvals, outdated certifications, or gaps between the tender’s stated rules and our compiled response file.
Core expectations
• Cloud-native stack: microservices, Docker containers, CI/CD pipeline, and infrastructure-as-code so we can deploy painlessly on AWS or GCP.
• Modern language tooling is up to you—Python, Node.js, or Go—provided the NLP pipeline remains transparent and easy to extend.
• Model layer should leverage reputable LLMs (OpenAI, Hugging Face, or similar) while allowing us to swap providers if policy or cost changes.
• Dashboard must support full-text search, filter by deadline, one-click export of compliance packs, and an audit trail of every action.
• Security: role-based access, encrypted data at rest, and logging that meets ISO 27001 style traceability.
Deliverables
1. Source code in a private Git repository with clear README and setup scripts.
2. Container images and IaC templates.
3. Initial trained or fine-tuned NLP models.
4. Working demo with at least three recent DGCP Technology tenders, showing the full flow from ingestion to compliance pack and audit report.
5. Brief hand-off session plus 30-day post-delivery bug fix window.
If you have built tender, RFP, or contract-analysis tools before, that’s a big plus. Please outline your proposed tech stack, any similar projects, and the timeline you’d need for a first functional release.