Windows GUI
Accepts annual reports or ZIP batches and runs text processing with sentence-level sentiment analysis.
01 / Paid client engagement · May–Aug 2025
From one company's report to
a batch tool the client can run.
Packaged topic filtering and sentence-level sentiment analysis into a Windows application, then extended batch processing and exports around the client's actual workflow.
- My role
- Requirements analysis, design, development, and handoff
- Deliverables
- Windows GUI, structured exports, and an operating guide
- Outcome
- Accepted and paid for after client use
The problem: useful results need a workflow people can run
The client needed to find topic-specific statements and identify sentiment across listed companies' annual reports. Manual review was slow, and a nontechnical user could not run model scripts directly. The client specified the goal, leaving the implementation open.
I broke the task into text processing, topic filtering, sentence-level classification, validation, and structured export. The tool supports investment research and human review.
The approach: complete one workflow, then expand from feedback
Start with one company
Connected text processing through export into a working end-to-end flow before adding batch support.
Expand to batch inputs
Added multiple annual reports and ZIP inputs, then adjusted XLSX fields and aggregation rules to support the client's follow-up analysis.
Deliver for the client's environment
Wrapped BERT-based sentiment analysis in a Windows GUI. Repackaged dependencies to address missing libraries, startup failures, slow batches, and connectivity issues, with a demonstration and operating guide.
Deliverables: processing, outputs, and handoff
Excel / CSV outputs
Produces structured data for further analysis, with export fields adjusted to the client's requirements.
Demonstration and operating guide
Documents setup and operation so the client can use the tool in their own environment.
This case is anonymized. Client materials, software screenshots, and acceptance correspondence are not public. Deliverables are described in text; illustrations are not presented as runtime evidence.
Validation: can the client process and export?
The client followed the guide to complete batch processing and structured exports, then accepted the delivery and paid.
This supports the claim of a usable delivery. No quantitative sentiment accuracy, time savings, or investment returns are established here.
The tool provides information for human analysis; investment decisions remain with the user.