Build AI Trade Intelligence Platform
Budget: $30 – $250 AUD
I need a fast-turnaround prototype of an AI-assisted trade-intelligence platform focused exclusively on competitor analysis. All development will happen inside a shared Replit workspace so we can collaborate and see progress in real time.
Core scope
• Ingest and normalise data from two primary sources: publicly available financial reports and curated market-research reports.
• Apply machine learning, natural language processing, and predictive analytics to extract competitor-specific insights—growth trajectories, market share shifts, pricing moves, and other red-flag signals.
• Present results through a simple web dashboard or REST API (whichever you can spin up quickest in Replit) with clear visual or JSON outputs that I can pipe into my own reporting stack.
Deliverables
1. Replit project link with fully runnable code and README.
2. Data-ingestion module covering both source types, including parsers, cleaning scripts, and test data.
3. ML/NLP models that tag, score, and forecast competitor behaviours, saved with reproducible training notebooks or scripts.
4. Lightweight UI or documented endpoints that surface the insights.
5. Brief hand-off session (recorded or live) walking me through deployment and extending the models.
Acceptance criteria
• End-to-end run completes inside Replit without local dependencies.
• At least one sample financial report and one market-research report processed, with the platform returning structured competitor metrics and a one-paragraph narrative summary per competitor.
• Code is cleanly commented so I can iterate later.
If you’re comfortable moving fast in Replit and have solid ML/NLP chops, jump in—I’m ready to review and merge code as soon as you push it.
Core scope
• Ingest and normalise data from two primary sources: publicly available financial reports and curated market-research reports.
• Apply machine learning, natural language processing, and predictive analytics to extract competitor-specific insights—growth trajectories, market share shifts, pricing moves, and other red-flag signals.
• Present results through a simple web dashboard or REST API (whichever you can spin up quickest in Replit) with clear visual or JSON outputs that I can pipe into my own reporting stack.
Deliverables
1. Replit project link with fully runnable code and README.
2. Data-ingestion module covering both source types, including parsers, cleaning scripts, and test data.
3. ML/NLP models that tag, score, and forecast competitor behaviours, saved with reproducible training notebooks or scripts.
4. Lightweight UI or documented endpoints that surface the insights.
5. Brief hand-off session (recorded or live) walking me through deployment and extending the models.
Acceptance criteria
• End-to-end run completes inside Replit without local dependencies.
• At least one sample financial report and one market-research report processed, with the platform returning structured competitor metrics and a one-paragraph narrative summary per competitor.
• Code is cleanly commented so I can iterate later.
If you’re comfortable moving fast in Replit and have solid ML/NLP chops, jump in—I’m ready to review and merge code as soon as you push it.