Vibe Coding a Non-public AI Monetary Analyst with Python and Native LLMs

Picture by Writer

# Introduction

 
Final month, I discovered myself gazing my financial institution assertion, making an attempt to determine the place my cash was truly going. Spreadsheets felt cumbersome. Present apps are like black bins, and the worst half is that they demand I add my delicate monetary information to a cloud server. I needed one thing completely different. I needed an AI information analyst that might analyze my spending, spot uncommon transactions, and provides me clear insights — all whereas retaining my information 100% native. So, I constructed one.

What began as a weekend venture was a deep dive into real-world information preprocessing, sensible machine studying, and the ability of native giant language fashions (LLMs). On this article, I’ll stroll you thru how I created an AI-powered monetary evaluation app utilizing Python with “Vibe Coding.” Alongside the best way, you’ll study many sensible ideas that apply to any information science venture, whether or not you might be analyzing gross sales logs, sensor information, or buyer suggestions.

By the top, you’ll perceive:

• Find out how to construct a strong information preprocessing pipeline that handles messy, real-world CSV recordsdata
• How to decide on and implement machine studying fashions when you could have restricted coaching information
• Find out how to design interactive visualizations that really reply person questions
• Find out how to combine an area LLM for producing natural-language insights with out sacrificing privateness

The entire supply code is obtainable on GitHub. Be happy to fork it, lengthen it, or use it as a place to begin to your personal AI information analyst.

Fig. 1: App dashboard displaying spending breakdown and AI insights | Picture by Writer

# The Drawback: Why I Constructed This

 
Most private finance apps share a elementary flaw: your information leaves your management. You add financial institution statements to companies that retailer, course of, and probably monetize your data. I needed a instrument that:

1. Let me add and analyze information immediately
2. Processed every thing domestically — no cloud, no information leaks
3. Offered AI-powered insights, not simply static charts

This venture grew to become my automobile for studying a number of ideas that each information scientist ought to know, like dealing with inconsistent information codecs, deciding on algorithms that work with small datasets, and constructing privacy-preserving AI options.

# Challenge Structure

 
Earlier than diving into code, here’s a venture construction displaying how the items match collectively:

venture/   
  ├── app.py              # Foremost Streamlit app
  ├── config.py           # Settings (classes, Ollama config)
  ├── preprocessing.py    # Auto-detect CSV codecs, normalize information
  ├── ml_models.py        # Transaction classifier + Isolation Forest anomaly detector
  ├── visualizations.py   # Plotly charts (pie, bar, timeline, heatmap)
  ├── llm_integration.py  # Ollama streaming integration
  ├── necessities.txt    # Dependencies
  ├── README.md           # Documentation with "deep dive" classes
  └── sample_data/
    ├── sample_bank_statement.csv
    └── sample_bank_format_2.csv

We are going to take a look at constructing every layer step-by-step.

# Step 1: Constructing a Sturdy Information Preprocessing Pipeline

 
The primary lesson I discovered was that real-world information is messy. Completely different banks export CSVs in utterly completely different codecs. Chase Financial institution makes use of “Transaction Date” and “Quantity.” Financial institution of America makes use of “Date,” “Payee,” and separate “Debit”https://www.kdnuggets.com/”Credit score” columns. Moniepoint and OPay every have their very own kinds.

A preprocessing pipeline should deal with these variations routinely.

// Auto-Detecting Column Mappings

I constructed a pattern-matching system that identifies columns no matter naming conventions. Utilizing common expressions, we will map unclear column names to plain fields.

import re

COLUMN_PATTERNS = {
    "date": [r"date", r"trans.*date", r"posting.*date"],
    "description": [r"description", r"memo", r"payee", r"merchant"],
    "quantity": [r"^amount$", r"transaction.*amount"],
    "debit": [r"debit", r"withdrawal", r"expense"],
    "credit score": [r"credit", r"deposit", r"income"],
}

def detect_column_mapping(df):
    mapping = {}
    for area, patterns in COLUMN_PATTERNS.objects():
        for col in df.columns:
            for sample in patterns:
                if re.search(sample, col.decrease()):
                    mapping[field] = col
                    break
    return mapping

The important thing perception: design for variations, not particular codecs. This method works for any CSV that makes use of frequent monetary phrases.

// Normalizing to a Customary Schema

As soon as columns are detected, we normalize every thing right into a constant construction. For instance, banks that break up debits and credit have to be mixed right into a single quantity column (damaging for bills, optimistic for revenue):

if "debit" in mapping and "credit score" in mapping:
    debit = df[mapping["debit"]].apply(parse_amount).abs() * -1
    credit score = df[mapping["credit"]].apply(parse_amount).abs()
    normalized["amount"] = credit score + debit

Key takeaway: Normalize your information as quickly as potential. It simplifies each following operation, like characteristic engineering, machine studying modeling, and visualization.

Fig 2: The preprocessing report exhibits what the pipeline detected, giving customers transparency | Picture by Writer

# Step 2: Selecting Machine Studying Fashions for Restricted Information

 
The second main problem is proscribed coaching information. Customers add their very own statements, and there’s no huge labeled dataset to coach a deep studying mannequin. We want algorithms that work nicely with small samples and will be augmented with easy guidelines.

// Transaction Classification: A Hybrid Strategy

As a substitute of pure machine studying, I constructed a hybrid system:

1. Rule-based matching for assured instances (e.g., key phrases like “WALMART” → groceries)
2. Sample-based fallback for ambiguous transactions

SPENDING_CATEGORIES = {
    "groceries": ["walmart", "costco", "whole foods", "kroger"],
    "eating": ["restaurant", "starbucks", "mcdonald", "doordash"],
    "transportation": ["uber", "lyft", "shell", "chevron", "gas"],
    # ... extra classes
}

def classify_transaction(description, quantity):
    for class, key phrases in SPENDING_CATEGORIES.objects():
        if any(kw in description.decrease() for kw in key phrases):
            return class
    return "revenue" if quantity > 0 else "different"

This method works instantly with none coaching information, and it’s simple for customers to know and customise.

// Anomaly Detection: Why Isolation Forest?

For detecting uncommon spending, I wanted an algorithm that might:

1. Work with small datasets (in contrast to deep studying)
2. Make no assumptions about information distribution (in contrast to statistical strategies like Z-score alone)
3. Present quick predictions for an interactive UI

Isolation Forest from scikit-learn ticked all of the bins. It isolates anomalies by randomly partitioning the information. Anomalies are few and completely different, in order that they require fewer splits to isolate.

from sklearn.ensemble import IsolationForest

detector = IsolationForest(
    contamination=0.05,  # Count on ~5% anomalies
    random_state=42
)
detector.match(options)
predictions = detector.predict(options)  # -1 = anomaly

I additionally mixed this with easy Z-score checks to catch apparent outliers. A Z-score describes the place of a uncooked rating by way of its distance from the imply, measured in normal deviations:
[
z = frac{x – mu}{sigma}
]
The mixed method catches extra anomalies than both technique alone.

Key takeaway: Generally easy, well-chosen algorithms outperform advanced ones, particularly when you could have restricted information.

Fig 3: The anomaly detector flags uncommon transactions, which stand out within the timeline | Picture by Writer

# Step 3: Designing Visualizations That Reply Questions

 
Visualizations ought to reply questions, not simply present information. I used Plotly for interactive charts as a result of it permits customers to discover the information themselves. Listed below are the design rules I adopted:

1. Constant coloration coding: Pink for bills, inexperienced for revenue
2. Context by means of comparability: Present revenue vs. bills aspect by aspect
3. Progressive disclosure: Present a abstract first, then let customers drill down

For instance, the spending breakdown makes use of a donut chart with a gap within the center for a cleaner look:

import plotly.categorical as px

fig = px.pie(
    category_totals,
    values="Quantity",
    names="Class",
    gap=0.4,
    color_discrete_map=CATEGORY_COLORS
)

Streamlit makes it simple so as to add these charts with st.plotly_chart() and construct a responsive dashboard.

Fig 4: A number of chart varieties give customers completely different views on the identical information | Picture by Writer

# Step 4: Integrating a Native Massive Language Mannequin for Pure Language Insights

 
The ultimate piece was producing human-readable insights. I selected to combine Ollama, a instrument for operating LLMs domestically. Why native as a substitute of calling OpenAI or Claude?

1. Privateness: Financial institution information by no means leaves the machine
2. Price: Limitless queries, zero API charges
3. Velocity: No community latency (although technology nonetheless takes a number of seconds)

// Streaming for Higher Person Expertise

LLMs can take a number of seconds to generate a response. Streamlit exhibits tokens as they arrive, making the wait really feel shorter. Right here is a straightforward implementation utilizing requests with streaming:

import requests
import json

def generate(self, immediate):
    response = requests.publish(
        f"{self.base_url}/api/generate",
        json={"mannequin": "llama3.2", "immediate": immediate, "stream": True},
        stream=True
    )
    for line in response.iter_lines():
        if line:
            information = json.hundreds(line)
            yield information.get("response", "")

In Streamlit, you’ll be able to show this with st.write_stream().

st.write_stream(llm.get_overall_insights(df))

// Immediate Engineering for Monetary Information

The important thing to helpful LLM output is a structured immediate that features precise information. For instance:

immediate = f"""Analyze this monetary abstract:
- Whole Earnings: ${revenue:,.2f}
- Whole Bills: ${bills:,.2f}
- High Class: {top_category}
- Largest Anomaly: {anomaly_desc}

Present 2-3 actionable suggestions primarily based on this information."""

This provides the mannequin concrete numbers to work with, resulting in extra related insights.

Fig 5: The add interface is easy; select a CSV and let the AI do the remainder | Picture by Writer

// Working the Utility

Getting began is simple. You have to Python put in, then run:

pip set up -r necessities.txt

# Optionally available, for AI insights
ollama pull llama3.2

streamlit run app.py

Add any financial institution CSV (the app auto-detects the format), and inside seconds, you will note a dashboard with categorized transactions, anomalies, and AI-generated insights.

# Conclusion

 
This venture taught me that constructing one thing useful is only the start. The actual studying occurred after I requested why every bit works:

• Why auto-detect columns? As a result of real-world information doesn’t comply with your schema. Constructing a versatile pipeline saves hours of guide cleanup.
• Why Isolation Forest? As a result of small datasets want algorithms designed for them. You don’t all the time want deep studying.
• Why native LLMs? As a result of privateness and value matter in manufacturing. Working fashions domestically is now sensible and highly effective.

These classes apply far past private finance, whether or not you might be analyzing gross sales information, server logs, or scientific measurements. The identical rules of strong preprocessing, pragmatic modeling, and privacy-aware AI will serve you in any information venture.

The entire supply code is obtainable on GitHub. Fork it, lengthen it, and make it your personal. When you construct one thing cool with it, I might love to listen to about it.

// References

Shittu Olumide is a software program engineer and technical author keen about leveraging cutting-edge applied sciences to craft compelling narratives, with a eager eye for element and a knack for simplifying advanced ideas. You can too discover Shittu on Twitter.