SEC 10-K Analysis & App Development
Budget: ₹600 – ₹3,000 INR
Task 1.1: Download Data from the SEC-EDGAR
The first part of the task is to download some SEC 10-K filings of public firms. These
documents will be analyzed in the next part of the task.
• First, select 2-3 companies or tickers of your choice.
• Second, for a given company ticker from a user, you need to download a 10-K filing
for each year from the SEC website for the period 1995 through 2023 (resulting in
approximately 29 10-K documents, it can be less depending on when the company IPOed).
• Note that the task is to write a script or program to do this automatically, and not to
download all files manually.
• You can use the sec-edgar-downloader (https://sec-edgar-downloader.readthedocs.
io/en/latest/) package in Python for ease or other similar packages.
Task 1.2: Text Analysis
• Note: You have to submit ONLY
– The program is used to merge data, clean data, perform analysis, and generate
visualization.
The main task for this part is to conduct text analysis using LLM API available for free.
• There are a few services that provide LLM inference API along with some free credits
to use. You don’t need to use any paid service.
• Use LLM inference API to gather some information or generate some insights from the
10K filings for a given company.
• note: we are not giving instructions on a specific ”insight”. The judgment calls you need
to make to decide on a ”good insight” is part of the assignment scoring. The assignment
is deliberately vague as we want you to make some good judgment calls.
• give 1-2 line explanation on why a user would care about this insight.
• Construct a visualization from the generated insight/information.
Task 2: Construct and Deploy Simple App
• Using the Task 1 code as the backend, create a simple app that takes the company ticker
as input and displays some visualization.
PS: Mention the insights you would use for task 1.2 in the proposals
The first part of the task is to download some SEC 10-K filings of public firms. These
documents will be analyzed in the next part of the task.
• First, select 2-3 companies or tickers of your choice.
• Second, for a given company ticker from a user, you need to download a 10-K filing
for each year from the SEC website for the period 1995 through 2023 (resulting in
approximately 29 10-K documents, it can be less depending on when the company IPOed).
• Note that the task is to write a script or program to do this automatically, and not to
download all files manually.
• You can use the sec-edgar-downloader (https://sec-edgar-downloader.readthedocs.
io/en/latest/) package in Python for ease or other similar packages.
Task 1.2: Text Analysis
• Note: You have to submit ONLY
– The program is used to merge data, clean data, perform analysis, and generate
visualization.
The main task for this part is to conduct text analysis using LLM API available for free.
• There are a few services that provide LLM inference API along with some free credits
to use. You don’t need to use any paid service.
• Use LLM inference API to gather some information or generate some insights from the
10K filings for a given company.
• note: we are not giving instructions on a specific ”insight”. The judgment calls you need
to make to decide on a ”good insight” is part of the assignment scoring. The assignment
is deliberately vague as we want you to make some good judgment calls.
• give 1-2 line explanation on why a user would care about this insight.
• Construct a visualization from the generated insight/information.
Task 2: Construct and Deploy Simple App
• Using the Task 1 code as the backend, create a simple app that takes the company ticker
as input and displays some visualization.
PS: Mention the insights you would use for task 1.2 in the proposals