Project: Cognito

By
Debiprasad Mohapatra
Vashu Saroha
Problem Statement
Brief Descroption of the Current Problem

In a global business landscape that is volatile,

uncertain, and highly data-driven, executives,
entrepreneurs, and financial strategists need real-
time access to robust, multi-dimensional data on
macroeconomic indicators. These indicators, such
as Gross Domestic Product (GDP), Consumer Price
Index (CPI) inflation rates, employment rates, and
financial indices like Nifty 50, are vital for making
informed strategic decisions. Without effective
access to historical data, trends, and predictive
analytics, businesses risk falling behind due to
untimely, uninformed, or reactive decision-
making.

1
Challenges
Below mentions the challenges faced in economic data analysis

Economic data is fragmented across sources i.e government

databases, financial news, proprietary platforms with
diverse formats (CSV, JSON, PDF) require extensive
processing. Customized comparisons, like correlating GDP
growth (5-7% annually) with inflation rates (typically 3-6%),
are often unavailable. Building analyses (e.g., regression or
five-year forecasts) demands costly expertise in
econometrics. While 10+ years of data is crucial for spotting
trends, many platforms restrict access or require
subscriptions upwards of $1,000 annually. Additionally,
integrating proprietary data (e.g., CSV, XLSX) with
macroeconomic indicators for targeted insights such as
inflation’s effect on Q1-Q3 retail sales remains challenging,
limiting responsive decision-making.

2
Solution
Below we propose a solution suited to cater to the current problem

Cognito, a “strategy consultant on demand” is a

centralized platform for decision-makers needing economic
insights. It provides real-time access to key indicators like
GDP, CPI, employment, exchange rates, and stock indices,
enabling cross-indicator analysis and comparison. Offering
10-year historical data and five-year forecasts, Cognito helps
users anticipate economic trends. Customizable, it supports
data uploads (XLSX, CSV, PDF, TXT) and prompt-based
queries (e.g., “Analyze inflation’s impact on Nifty 50”).
Dynamic visuals—trend charts, line graphs, and correlation
plots—make complex data relationships actionable for
strategic planning.

3
Feasibility Study
Below we propose a Feasibility Study suited to cater to the current problem

The feasibility of the web application includes technical,

operational, financial, and legal aspects. The tech stack
involves Flask for the backend, Vanila JS & Bootstrap for the
frontend, MySQL for database, and data analysis with
Pandas and Scikit-learn. Cloud hosting ensures scalability.
User authentication will use OAuth, and data privacy will
comply with GDPR , CCPA and DPDP. Financially, the
platform will adopt a subscription or freemium model. Legal
compliance is essential, especially for data privacy and API
licensing. Risks include data security, accuracy, and
scalability, mitigated through strong encryption, reliable
data sources, and cloud solutions with auto-scaling.

4
Feasibility Study
Below we propose a DFD suited to cater to the current problem

The Data Flow Diagram (DFD) for the Cognito web application illustrates the flow of data within the system,
focusing on key components: Users, File Upload, Analysis, and Macro Indicators. Users interact with the system to
upload data files, which are stored in the system’s storage and linked to the database. The Analysis module allows
users to input prompts, analyze the uploaded data, and obtain results. Additionally, the Macro Indicators module
provides access to economic indicators like GDP and inflation, which can be compared or analyzed alongside user
data. The system generates and returns insights, forecasts, and comparative analyses to the user interface.

5
Use Cases
Below we mention the usescases of this model.

The Cognito platform enables users to register, log in, and

securely upload data files (CSV, XLSX, etc.) for analysis.
Users can specify prompts for analyses like forecasting or
comparison, and the system returns insights based on
uploaded data and macroeconomic indicators (e.g., GDP,
inflation, Nifty 50 trends). A comparison tool allows users
to observe correlations, while audit logging tracks user
actions for security. The platform generates visualizations,
reports, and downloadable summaries for decision-
making. These key features support a streamlined, data-
driven process for users seeking strategic business
insights from economic and custom datasets.

6
Graphical User Interface
Below we mention the GUI of this model.

The Cognito platform offers a user-friendly experience

with features across key pages. The Landing Page
highlights Cognito’s value and CTAs for sign-up/login.
Users securely access the platform via User Registration
and Login. The Dashboard centralizes navigation, data
uploads, and updates. The Data Upload Page supports
various file types and tracks upload history, while the
Analysis Page allows users to enter prompts, select
analysis types, and view past analyses. The Indicators Page
displays interactive economic indicators, and the Analysis
Results Page provides downloadable insights. Settings lets
users manage profiles, preferences, and billing for paid
plans.

7
Design and
Implementation
ER Diagram
Below we mention the ER Diagram of this model.

The database schema consists of seven interconnected tables

designed to manage and track user activities, data files, and analyses
on the Cognito platform. The Users table stores user information,
including credentials and roles. The DataFiles table records uploaded
files, linking them to specific users. ValidationLogs track the validation
process of each file, capturing status and details. AnalysisRequests
logs the analysis requests made by users, specifying the type of
analysis and its current status. The AnalysisResults table stores the
outcomes of completed analyses. MacroeconomicIndicators holds
data on key economic indicators, including values and forecasts.
Finally, the AuditLogs table monitors and records user actions for
auditing and security purposes, ensuring transparency and
accountability across the platform. Together, these tables provide a
comprehensive data structure for efficient management of user data,
file uploads, and analytical processes.
9
Database Schema
Below we mention the Database Schema of this model.

The database schema comprises seven interconnected tables designed for a data analysis system. The Users
table forms the foundation, storing user credentials and roles. DataFiles tracks uploaded files, while
ValidationLogs monitors file validation status. AnalysisRequests and AnalysisResults handle data analysis
workflows, storing request details and outcomes respectively. MacroeconomicIndicators maintains economic
data independently. Finally, AuditLogs tracks system activities for security. The schema implements referential
integrity through foreign key relationships, with cascading deletes ensuring data consistency. Primary keys use
auto-incrementing integers, and timestamps track creation and modification times across tables.
10
Archetectural Design
Below we mention the Archetectural Design of this model.

The architectural design presents a modern

microservices-based system structured in four distinct
layers. The Client Layer offers web and mobile
interfaces, while the API Gateway handles
authentication, routing, and rate limiting. The
Application Layer comprises core services (User, Data,
Validation, and Analysis) and support services (Audit
Logging and Metrics). The Data Layer manages
storage through various databases: a relational
database for main data, file storage for documents,
Redis for caching, InfluxDB for time series data, and
Elasticsearch for search capabilities. This layered
architecture ensures scalability, maintainability, and
secure data processing.

11
Project Management
Gantt Chart
Below we mention the Project Gantt of this model.

The architectural design presents a modern

microservices-based system structured in four
distinct layers. The Client Layer offers web and
mobile interfaces, while the API Gateway handles
authentication, routing, and rate limiting. The
Application Layer comprises core services (User,
Data, Validation, and Analysis) and support services
(Audit Logging and Metrics). The Data Layer
manages storage through various databases: a
relational database for main data, file storage for
documents, Redis for caching, InfluxDB for time
series data, and Elasticsearch for search
capabilities. This layered architecture ensures
scalability, maintainability, and secure data
processing.
13
Function Point Analysis
Below we mention the Complexity Weighting of this model.

Complexity Weighting

14
Function Point Analysis
Below we mention the VFA & FA of this model.

VAFs

The Function Point Analysis includes two components: the

Unadjusted Function Points (UFP) table showing different
function types (EI, EO, EQ, ILF, EIF) with complexity ratings (Low,
Medium, High) and Project Rating, totaling 61 points; and the
General System Characteristics (GSC) table evaluating 14
technical factors rated from 1-3, summing to 34 points. These
tables help assess software size and complexity.

FP={UAF*0.65+(0.01*Total VAF}
FP={61*0.65+(0.01*Total VAFs)}

FP=39.99~ 40

15
Risk Management and Analysis
Below we mention the potential Risks associated with this model.

Risk Categories
Technical risks in the project encompass several key
categories. Database risks include configuration errors, data
corruption, or inefficiencies during transactions or backups.
Security risks involve vulnerabilities in data handling,
encryption, and authentication mechanisms, potentially
enabling unauthorized access. Code quality and testing risks
arise from logical flaws, unoptimized code, and insufficient
testing, leading to bugs and performance issues. Scalability
and performance risks threaten the system’s ability to
handle increased traffic, causing slow performance or
crashes. Dependency risks stem from vulnerabilities in third-
party libraries. Additionally, deployment risks and API risks
can result in downtime, insecure data exposure, or degraded
system performance.
16
Risk Management and Analysis
Below we mention the Impact Analysis of this model.

Impact Analysis
The project faces several technical risks, each with
varying levels of impact and likelihood. Database risks
have a high impact and medium likelihood, making
them a high priority. Security risks are deemed critical
due to their high likelihood and severe consequences.
Code quality and testing risks are medium in both
impact and likelihood, resulting in a medium priority.
Scalability and performance risks are critical, with a
high likelihood of occurrence. Dependency risks and
API risks both carry high impact and medium
likelihood, making them high priorities. Deployment
risks, though high in impact, are less likely, ranking as
medium priority.

17
Risk Management and Analysis
Below we mention the Mitigation Strategies.

Mitigation Strategies
To mitigate project risks, secure database credentials, automate
connection testing, and use ACID-compliant databases with
regular backups and replication. Enhance security by enforcing
encryption, HTTPS, OAuth 2.0, and multi-factor authentication.
Ensure code quality with static analysis, peer reviews, and
automated testing frameworks like Pytest and Selenium.
Optimize scalability and performance using load balancers,
caching, and CDNs while profiling for bottlenecks. Manage
dependencies by monitoring vulnerabilities, locking versions,
and automating updates with tools like Dependabot.
Standardize deployments using Docker, CI/CD pipelines, and
staging validations. Protect APIs with authentication, input
validation, and performance monitoring through tools like
Postman and Swagger.
18
Testing
Testing
Below we mention the Various Test Cases of this model.

Test Cases

21
Testing
Below we mention the Flow Graph of this model.

Flow Graph

The flow graph illustrates the core system workflow, beginning

with user authentication through the login process. Upon
successful login, users access the dashboard, where they can
either upload files or view reports. File processing involves
validation checks - valid files proceed to analysis while invalid
ones trigger error handling. Completed analyses generate
results that feed into the reporting system. Users can view these
reports and export them for download. The flow is organized
into three main subsystems: Authentication, Data Processing,
and Reporting, ensuring a logical progression from data input to
final output.

22
Testing
Below we mention the Cyclometric Complexity of this model.

Cyclometric Complexity

The cyclomatic complexity analysis

reveals that the Data Processing and
Analysis Engine modules have the
highest complexity (7 and 9
respectively) due to multiple decision
points and data transformation paths.
The Authentication and File Upload
modules show moderate complexity (4-
5), while Report Generation and Export
modules maintain low complexity (3).
This suggests focusing testing efforts
on the high-complexity modules and
potential refactoring of the Analysis
Engine to reduce complexity.
23
Thank You