Budget: 800 USD Deadline: 15 days

I can help with the development of a local Python tool for reconciling invoices, bank statements, card expenses, and accountant files. I have experience with Python, Excel processing, pandas/openpyxl, API integrations, automation, and data structuring. I would like to clarify:
1. In what format and structure will the bank statements, card files, and accountant file be?
2. How many emails/attachments need to be processed in one period?
3. In what language are the invoices usually: Hebrew, English, Russian, or multiple languages?
4. Is the Claude API already available, or does it also need to be connected and configured?
5. Is it necessary to store the downloaded invoices locally after processing?
6. What rules define "recognized expenses" — are they already established or need to be described separately?
7. What is the acceptable deviation for dates and amounts during reconciliation?
8. Is only a CLI launch needed, or is it preferable to create a simple desktop interface?
9. Are there examples of Excel files and invoices for testing?
10. What timelines are you considering for implementation? Expand

Budget: 200 USD Deadline: 1 day

Hello!

We can create a local Python tool for such reconciliation for you.

1. Can you send examples of Excel files and invoices to check the data structure?
2. Which part should we implement first: collection from Gmail, reconciliation, or the final Excel report?

—
About us

We are dZENcode – a full-cycle IT development company: from design and programming to integrations and post-release support of digital solutions. We create projects from scratch, as well as connect to existing solutions that need refinement, development, or technical support.

You can find detailed information about our services and rates on our official website: Freelancehunt

Take a look – after that we can discuss the details and agree on the next step.

⚠️ After clarifying all the details, we will determine the scope of work, the suitable format of cooperation: task-based, outsourcing, or outstaffing – and the final cost.

With us, projects are guaranteed to reach release:
• 10+ years providing IT services;
• 90+ in-house specialists;
• 250+ public reviews since 2015;
• We support products under SLA after launch;
• We work under NDA and contract with the company! Expand

Budget: 1500 USD Deadline: 7 days

Good day, Grisha!

I reviewed your project and propose to develop a local application in Python for reconciling accounts with the bank, cards, and the accountant. The main stages include integration with the Gmail API for downloading invoices, recognizing information using OCR, and the logic for matching data. I am confident I can implement all the necessary functions to automate this process.

I have experience working with APIs and OCR technologies, which will help ensure the accuracy of invoice reconciliation and avoid accounting errors. I am ready to create a tool that will simplify your work and increase efficiency.

To better understand your goals, please clarify:
- Launch deadlines: do you need an MVP or is there some time buffer?
- Is there a clear technical specification or vision for the tool?
- What volume of data is planned for processing: how many invoices and transactions?
- What content for data recognition do you have: are the PDF files and images ready or will they need to be uploaded?

These details will help me propose the optimal solution.

I would also note that the development process, timelines, and costs may be influenced by the following factors:
1. Whether we use ready-made libraries or develop unique solutions.
2. Whether you will provide the content for processing or if I need to gather it.
3. Whether a custom interface is needed or if a standard one is sufficient.
4. Whether there is a detailed technical specification or if we will work based on general directions.
5. What volume of data is planned for processing and how many types of transactions need to be reconciled.

At the start, it is important to form a general vision of the final result. Let's discuss the functions and features that you consider important, as well as examples that inspire you. This will help clearly formulate the technical specification.

I look forward to your response to schedule a time for communication and discuss the project details. Expand

Budget: 2500 USD Deadline: 20 days

Grisha, the task is clear: a local Python tool for regular reconciliation of invoices, bank statements, cards, and accountant data. I can build a solution with the Gmail API, PDF/photo parsing, OCR with Hebrew support, Excel upload, and flexible matching logic by amount, date, and counterparty. I will also create a report in one Excel file highlighting discrepancies and a clear CLI instruction. I have experience in Python automation and integrations, ready to discuss the structure of your files and start quickly.

Budget: 700 USD Deadline: 21 days

Hello! The specifications are clear, and the project is definitely within my specialization. I will create a Python-CLI that performs all 4 checks and produces one Excel file with multiple sheets and highlighting of problematic rows.

How I will do it:

Email collection: Gmail API (OAuth), searching for emails by period, extracting attachments (PDF + images).
Recognition: PDF — text extraction (pdfplumber/PyMuPDF); images — I agree with you, for Hebrew, Claude API (vision) provides significantly better accuracy than Tesseract.
Excel: reading bank statements, card files, and accountant files using pandas/openpyxl.
Matching: amount + date (with tolerance) + counterparty (fuzzy name matching, rapidfuzz).
Checks A-D: income without invoice, expenses without invoice, reconciliation with the accountant's list (both ways), payroll vs bank.
Report: one .xlsx, a separate sheet for each check, discrepancies highlighted in color. Plus, instructions for running and setting up Gmail/Google Cloud.
Relevant experience: I have done integrations with Google API and generated structured Excel reports with validation and data matching (there is an example in my portfolio — a parser-aggregator with a 24-column report). I have worked extensively with PDFs/tables.

A few clarifications (can be based on your example files — I would be happy):

How to determine "recognized expenses" (by category in the statement or a separate directory)?
What tolerances are acceptable for amount and date during matching?
In what format is the payroll (PDF/Excel)? Expand

Budget: 350 USD Deadline: 7 days

Hello, Hryhoriy! My name is Oleksandr. I am an analyst with a master's background in business economics and an AI solutions developer. Automating financial accounting, matching statements, and building reconciliation algorithms is my niche.

Your technical specification is formulated as professionally as possible. I fully support the use of the Claude API (Anthropic model) instead of Tesseract — this will ensure 100% accuracy in recognizing Hebrew and numbers on receipt photos due to the model's powerful vision capabilities.

I propose a turnkey tool architecture that will fully address your task:

1. DATA COLLECTION AND PARSING (GMAIL + EXCEL + PDF):
— I will connect to the email via the official Gmail API and OAuth (using google-api-python-client). We will jointly set up the project in Google Cloud Console and generate credentials.json.
— The script will automatically download PDF invoices and receipt photos for the selected period. For PDFs, we will use pdfplumber/PyPDF, and for images — the Claude API, which will immediately convert the data into structured JSON (date, amount, counterparty).
— Using Pandas, we will read bank statements, credit card files, and the accountant's Excel files.

2. MATCHING AND VERIFICATION LOGIC:
— Since there is no single key, matching will be based on a combination of markers: [Date with an allowable deviation of 2-3 days for bank processing + Exact amount + Fuzzy comparison of counterparty names (Fuzzy Matching)].
— The algorithm will perform all necessary checks: the presence of invoices for all income, retention of receipts for recognized expenses, reconciliation with the accountant's list for tax/VAT, and verification of the amounts in the payroll against the bank's withdrawal.

3. RESULT AND SUBMISSION FORMAT:
— The program will have a simple and fast CLI interface (launch in the command line with 1 click without overloading web interfaces).
— The script will generate one local Excel file (using openpyxl) with separate sheets: "Income without issued invoice", "Expenses without saved receipt", "Invoices missing from the accountant", "Payroll reconciliation with the bank", etc. Discrepancies will be visually highlighted in color.
— I will provide the complete source code of the project, a brief instruction for launching, and a guide for creating credentials in Google Cloud Console.

The development cost is 350 USD. The deadline is 7 days. I am ready to start working through a secure Safe agreement. Please send examples of the structure of your Excel files for designing the matching logic! Expand

Budget: 2400 USD Deadline: 21 days

Regarding the timeline and budget - I would estimate the first working version of such a local tool at $2400 and 2-3 weeks, provided there are examples in Excel, PDF, and photos of receipts. The stack of Python, pandas, openpyxl, PyMuPDF, or pdfplumber, Gmail API, and OAuth is suitable. For photos in Hebrew, I would separately check the recognition quality through the Claude API, because here accuracy is more important than beautiful architecture =/

Look, there’s a nuance - in the accountant's file, there is no single account number, so the bank, cards, Gmail attachments, and accounting Excel need to be first brought to a common data model, and then matched by amount - date - counterparty with allowances and the reasons journal. !!Financial reconciliation should not be a black box!!

> We will retrieve attachments from Gmail for the period
> We will extract the date, amount, counterparty, and account number from PDF and photos
> We will normalize bank, card, and accounting Excel files
> We will create matching rules and separate checks for income, expenses, accountant data, and payroll
> The output will be an Excel report with sheets, highlighting problematic rows, and a brief explanation of the reasons

I need anonymized examples of bank statements, cards, the accountant's file, and 5-10 documents in Hebrew from you. For Gmail - either we create credentials in Google Cloud Console together as per the instructions, or you provide a ready access file through a secure channel of the exchange.

I would like to clarify 2 points
> Are card expenses already marked as recognized expenses, or should this classification also be included in the rules?
> Can documents be sent to the Claude API for recognition, or is a mode without external file transfer needed?

Similar works by Ingello
> AI and data processing automation - https://business.ingello.com/vorfahr
> Corporate accounting systems and business logic - https://business.ingello.com/platforma
> Processes, accounting, and matching rules - https://business.ingello.com/forma-bpm

FLH landing Ingello - https://systems-fl.ingello.com Expand

Budget: 200 USD Deadline: 7 days

In such a reconciliation, it is often not the code that breaks, but the format matching: banks, cards, and accounting Excel files may refer to the same payments differently.

I would create a local Python tool that extracts data from Gmail, reads Excel/PDF/images, standardizes amounts and dates into one format, and shows discrepancies in a separate report.

For such tasks, it’s better to immediately create an error log and a list of "not recognized" items, so the accountant doesn’t have to manually search for the problem across all files.

I can start today: first, I will build a working prototype for one reconciliation scenario, and after verification, I will refine the rest of the rules.

Please let me know in what format you prefer to receive the final report — a separate Excel file or several tabs in one file.

Budget: 200 USD Deadline: 7 days

Hello, Grisha! I did something similar in Python — parsed bank statements and PDF invoices, matched them by amount, date, and counterparty using fuzzy name matching. Hebrew in the photo is better processed through Claude Vision, you are right here — Tesseract makes many mistakes with it. One important point: the payment date in the bank usually does not match the invoice date, what date range should we set — a week, two? This affects how leniently we match the matches. I'm ready to take it on, the estimate is 7 days, I will provide the source files and a short instruction on how to run it.

Budget: 125 USD Deadline: 7 days

Hello! I have reviewed your technical assignment. The specifications are incredibly clear, and I fully understand the architecture of the tool and the logic of the report. I am ready to implement this project "turnkey" in Python. I have ready experience and developments for all key modules of your task:

Data and Excel handling: I have extensive experience in automating spreadsheets using pandas and openpyxl libraries. I can create a summary multi-page report with color highlighting of problematic rows without any issues.

Parsing and reading documents: I have implemented modules for extracting plain text from PDFs using PyMuPDF and integration with IMAP/Gmail API for automatic downloading of attachments.

AI integration (OCR on neural networks): I work with the integration of local and cloud AI models (OpenAI / Ollama API) for intelligent text recognition and strict parameters (date, amount, counterparty) from unclear photographs of receipts.

My implementation plan:

Creating a modular CLI program, where each stage (downloading, recognition, analysis, final Excel) will be zoned.

Complete logging of processes in the console, so you can see at which stage a discrepancy occurred.

Preparing a clear instruction for launching and configuring the .env file (Google credential files).

I would be happy to receive test examples of the structure of your Excel files and photos of invoices for conducting local tests of the matching logic.

I am ready to discuss details, timelines, and costs in the chat! Expand

Budget: 300 USD Deadline: 10 days

Greetings! This is a great task; automating routine reconciliation through Python is the most reliable way to save dozens of hours before reports. The matching logic is completely clear to me: I will implement a Python script (pandas + openpyxl) for parsing statements. For fuzzy matching of counterparties (when names differ slightly), we will use the thefuzz / RapidFuzz library with a similarity threshold setting. We will automate document collection from email using the official google-api-python-client (OAuth 2.0). Regarding OCR: I fully support the decision to use the Claude API (Sonnet models) instead of Tesseract — Claude recognizes document structures, dates, and amounts much more accurately, especially in Hebrew (right to left). In the end, we will create a neat Excel report with color coding of discrepancies using openpyxl. I am ready to discuss the structure of your files and get started!

Budget: 500 USD Deadline: 7 days

Good day!

I professionally develop scripts and automation tools in Python. I am ready to implement this local tool for invoice reconciliation for you. I have extensive experience working with the Gmail API (OAuth), pandas, openpyxl, data extraction using pdfplumber, and integrating the Claude API for accurate Hebrew OCR and fuzzy searching.

Message me privately, and we will clarify the details.

Budget: 600 USD Deadline: 3 days

Hello, I have extensive experience working with Google APIs and spreadsheets. I can easily complete your task in the Python programming language. If you want to receive quality work within a reasonable timeframe, please message me to discuss the details.

Budget: 1000 USD Deadline: 6 days

The task is clear: a local Python tool for reconciling invoices from Gmail with the bank, cards, and the accountant's file. These are exactly the things I'm interested in.

How I will implement it: Gmail API (OAuth) for downloading attachments, pdfplumber for extracting text from PDFs, Claude API for OCR of images with Hebrew (specifically Tesseract in Hebrew), pandas + openpyxl for reading Excel and generating the final report. Matching records by amount, date, and fuzzy matching of the counterparty's name (rapidfuzz). The result: one Excel file with color coding for all four checks.

A couple of clarifications: (1) Approximately how many rows are in the bank statement for the period and how many invoices on average over 2 months? (2) Is the accountant's Excel file in Hebrew or in Russian/English? (3) Are you ready to provide sample files for testing right away?

I am ready to start this week; I will name a fixed price after your answers.

Budget: 2100 USD Deadline: 14 days

Hello. My name is Alexey, and I represent a group of developers – NC-1. For over five years, we have been creating websites, mobile applications, online stores, ERP/CRM systems, and other e-commerce products.
In our team, we have a Python developer, a senior with the necessary experience and knowledge for you. The preliminary cost and timeline: 1,900 USD, 14 working days.
Please send examples of the structure of Excel files (bank, cards, accountant file) and examples of invoices (PDF and photo) for testing.
Best regards, Alexey M.

Budget: 2400 USD Deadline: 14 days

We already have a practically ready similar solution for document, spreadsheet, and financial operation reconciliation, which can be quickly adapted to your scheme and discussed here on the marketplace; I am available =)
The timeline is aimed at 10-14 working days for the first working tool with Gmail, Excel, PDF and photo recognition, matching by amount, date, and counterparty name, plus a final Excel with highlighted problematic rows.
For the budget for this stage, I would estimate 2400 USD, more accurately after reviewing 2-3 examples of bank files, cards, accountant files, and a couple of invoices in Hebrew.

- Important nuance regarding OCR
-- For Hebrew, I would not only use classic Tesseract; it’s better to create a recognition layer through a model with confidence checking and a manual log of disputed documents.
-- This reduces the risk of silent errors in VAT and reporting, and this is a place where automation should be more cautious than my internal accountant after the third coffee.

- Questions
-- In the bank and card Excel files, are the dates and amounts always in one format, or does each file have its own structure?
-- Regarding the payroll sheet, is it a separate PDF from the email or a separate line in the accounting file?

- Similar cases
-- https://business.ingello.com/vorfahr - close in AI data processing and automation of manual operations.
-- https://business.ingello.com/platforma - close in accounting logic, roles, data, and management processes.
-- https://business.ingello.com/forma-bpm - close in formalizing business processes and checks between sources.

Main landing page for the marketplace - https://systems-fl.ingello.com Expand

Budget: 1500 USD Deadline: 15 days

Hello. I have extensive experience with Python. I am ready to complete the task.

Budget: 400 USD Deadline: 7 days

Hello!

I carefully studied the technical specifications - the task is clear. I have experience with Python, Gmail API, PDF parsing (pdfplumber/PyMuPDF), processing Excel through pandas/openpyxl, and Claude API for OCR including text recognition in Hebrew.

What I will implement:
Integration of Gmail API with OAuth for collecting invoices
Extraction of text from PDF + Claude API for OCR of photos/images with support for Hebrew
Fuzzy matching logic based on amount, date, and contractor name
All 4 checks (income, expenses, accountant list, payroll)
Final Excel report with color highlighting of discrepancies
Instructions for setting up Gmail credentials

I am ready to start after receiving examples of Excel files and samples of invoices for testing. Expand

Budget: 560 USD Deadline: 10 days

Hello,

I can build this as a clean Python-based reconciliation tool that runs locally on demand and handles the full workflow end-to-end.

My approach would start with a modular pipeline: Gmail API ingestion (OAuth-based) to pull invoice attachments for a selected date range, followed by structured extraction using PDF parsing for digital invoices and OCR for images, with special handling for Hebrew text using a high-accuracy vision model when needed. All extracted data would be normalized into a consistent schema (date, amount, counterparty, reference metadata).

For the financial sources, I would build robust parsers for Excel bank statements, credit card exports, and the accountant’s report using pandas, with validation layers to standardize formats across different banks and card providers.

The reconciliation engine would then apply fuzzy matching logic across amount, date tolerance, and counterparty similarity, and generate clear discrepancy categories such as missing invoices, unmatched expenses, accountant mismatches, and payroll deviations.

The final output would be a structured Excel report with multiple sheets and conditional formatting to highlight issues clearly. I would use openpyxl for formatting and ensure the file is easy to interpret for accounting review.

The project would be delivered as a CLI tool with clear setup instructions, including Gmail OAuth setup, dependency installation, and a simple configuration file for file paths and matching rules.

Best, Hanh Expand

Budget: 650 USD Deadline: 12 days

Hi hope you are doing well

I can build a local Python reconciliation tool that you can run manually every one or two months to compare Gmail invoices bank statements credit card files and the accountant Excel file.

My approach will be to create a CLI tool that connects to Gmail through OAuth searches emails by date range downloads PDF and image attachments and extracts invoice data such as date amount counterparty name and invoice number when available. For PDFs I can use direct text extraction and for images I can support Hebrew OCR using Claude API or another reliable OCR option.

The tool will read bank card and accountant Excel files with pandas and openpyxl then match records using amount date tolerance and fuzzy name comparison.

The final Excel report will include income without issued invoice expenses without saved invoice invoices missing from accountant accountant records without invoice and payroll reconciliation. Problem rows will be highlighted clearly.

Deliverables include source code setup guide Gmail API credential instructions test run support and local Excel output.

Best regards
Youssef. Expand

Budget: 140 USD Deadline: 5 days

Hello! I am ready to complete your extensive project at the minimum market price. It will take about 5–7 days, as I need to carefully fine-tune the fuzzy matching of counterpart names and test the work on your examples. Regarding the price, I will align with your minimum budget of 6000 hryvnias, as I am currently building reviews on the platform. If this is relevant, please send examples of files in private messages, we will look at the structure and get started. We work through Safe.

Budget: 550 USD Deadline: 14 days

Hello! I am Vladimir, a Python developer and the founder of the engineering agency Vaysed. Your technical assignment is described in great detail and competently, which allows for an immediate understanding of the business logic of the future tool. The task of automating financial reconciliation and data parsing perfectly fits my technology stack, and I am ready to implement this local script turnkey. I have extensive practical experience working with various APIs, processing unstructured data, and automating complex reconciliation processes. Creating such a local application in Python will allow you to securely process financial documents without transferring confidential files to third-party servers.

The technical implementation of the project will begin with setting up secure authorization via OAuth to collect data through the Gmail API. The script will scan emails for a specified period, filter messages, and locally save attachments for further processing. You are absolutely right about using the Claude API; it is the perfect solution for accurately recognizing Hebrew and extracting structured data from receipts, which will return information in a convenient format. For regular PDF documents, I will use the pdfplumber library for direct text extraction, which will significantly save resources and speed up the work.

To match counterparties from different sources, I will apply fuzzy search algorithms; the script will compare records based on strict criteria of amount and date, as well as a soft criterion of company name similarity. All logic checks, including searching for income without invoices, expenses without receipts, reconciliation with the accountant, and payroll analysis, will be implemented through powerful pandas dataframes, ensuring high speed even with large volumes of data. The final result will be exported to a clean Excel file with sheet separation, where I will set up conditional formatting through openpyxl so that all discrepancies are automatically highlighted for quick visual control.

In the end, you will receive a fully ready-to-use product, including clean source code of the project with comments and a file with dependencies for quick installation. I will also prepare a detailed step-by-step guide on setting up the project in Google Cloud, obtaining keys, entering the Claude token, and running the script from the command line. You can familiarize yourself in detail with my engineering approach and examples of already implemented projects in my portfolio at the link Freelancehunt.

Budget: 400 USD Deadline: 7 days

Hello!

I have extensive experience in creating systems and applications in Python.
You can view examples of my work in my portfolio.
I am ready to complete the assigned task.

I suggest discussing the details in private messages.

Budget: 350 USD Deadline: 10 days

I understand the task: a local Python tool is needed that runs manually once every 1-2 months on your computer and checks four sources. Invoices from Gmail (PDF and photo receipts), bank statements in Excel, card expenses in Excel from several files, and the accountant's Excel. The output will be one Excel file with sheets: income without an issued invoice, expenses without a saved invoice, discrepancies with the accountant's list, and payroll reconciliation, highlighting problematic rows.

Here’s how I see the implementation. Gathering from email via Gmail API using OAuth, selecting emails for a period with attachments, downloading PDFs and images. Text is extracted directly from PDFs using pdfplumber or PyMuPDF. For photos and scans of receipts in Hebrew, I agree with you that it’s better to use Claude API with image support rather than classic Tesseract: the accuracy in Hebrew from photos is significantly higher, which is critical here. Excel is read using pandas and openpyxl.

I build the matching based on a combination of amount plus date with an acceptable deviation plus the counterparty's name through fuzzy matching to catch matches even when the name is written slightly differently. Since there is no unique invoice number in the accountant's file, I first bring all sources to a single data model, and then match with the reason log so that the reconciliation is not a black box and you can see why a row did not match.

This is my main area of expertise: Python, pandas, working with Gmail and email, Anthropic API in production for two years, PDF parsing. I will make the tool so that you can run it yourself with one command from the console and add new card files without code modifications. I will provide the source code, a launch instruction, and instructions for creating credentials in Google Cloud.

One question that affects the reliability of matching: what date range do we set between the invoice date and the payment date in the bank, one week or two? Also, are the receipt photos flat scans or phone photos taken at an angle? This affects how aggressively we set up the recognition.

Budget: 550 USD Deadline: 6 days

Hello! The task is clear, I have experience working with the entire specified stack: Python (Pandas/Openpyxl) for processing Excel, working with Google API, and integrating LLM for recognizing complex documents.

Here’s how I propose to implement it:

Gmail API: I will implement secure authorization via OAuth 2.0. The script will selectively download attachments for the specified period.

Document processing: For PDFs, I will use pdfplumber to extract text. For images (including Hebrew), I will integrate the Claude API, as you requested — this will provide significantly higher accuracy than the classic Tesseract.

Matching logic: I will use fuzzy matching algorithms to correctly match the names of counterparties, even if they differ slightly in the statement and invoice.

Result: I will create a summary Excel file highlighting discrepancies in color (styling via openpyxl), so you can immediately see the problematic items.

I have experience automating similar routine processes. I am ready to start immediately after receiving sample files to set up the matching logic. Expand

Budget: 200 USD Deadline: 7 days

Hello!

My name is Dmytro Vilyhotskyi, I am a full-stack developer with experience in developing Telegram Mini Apps, SaaS platforms, and automation systems. For the last 7 months, I have been actively building projects on Python/FastAPI + React, integrating complex APIs and data processing systems. This task perfectly fits my work stack.

I am ready to discuss the details and work on the project.

Budget: 200 USD Deadline: 3 days

I have created similar tools for small businesses where it was necessary to consolidate data from different sources and identify discrepancies. I see it like this: we parse exports from the bank, cards, and accounting into a unified format, match transactions by amount, date, and counterparty, and generate a report with discrepancies and unmatched lines. Most likely, a console script with a simple HTML report output will be sufficient, without unnecessary dependencies. Please let me know in what format the exports from the bank and accounting come, is it csv/xls or something custom? I am ready to discuss the details and estimate the timeline.

Budget: 100 USD Deadline: 1 day

Hello. I can implement such a tool in Python: downloading invoices from Gmail, processing PDFs and images, reading bank Excel files, matching records, and generating a final report.

I have worked with Python, Gmail API, Excel, PDF, and OCR. First, I need to look at sample files and invoices, as the recognition and matching logic depends on their structure. After reviewing them, I will be able to more accurately assess the timeline and cost.

Budget: 750 USD Deadline: 7 days

Hello, Grisha!

I read the technical specifications carefully - the task is clear. A Python script for reconciliation with the Gmail API, PDF parsing, and OCR through Claude Vision (for Hebrew in photos - the right choice, Tesseract is weaker here).

What I will implement:
- Gmail OAuth + searching and downloading attachments for a specified period
- PDF: pdfplumber/PyMuPDF; photo receipts: Claude API Vision with Hebrew support
- Reading bank statements and cards through pandas
- Fuzzy matching by amount + date (with a configurable window) + fuzzy comparison of counterparty names
- All 5 checks from the technical specifications (A-D + payroll)
- Final Excel with separate sheets and color highlighting of discrepancies
- CLI + instructions for launching and connecting to Gmail

One question: what is the acceptable deviation in dates between the invoice and the payment in the bank - 3 days or 7? This affects the matching logic.

I am ready to start after receiving sample files. Deadline is 7 days. Expand

Budget: 850 USD Deadline: 6 days

Good day! I develop in Python, have worked on similar projects with React/Node.js, and am ready for collaboration.

Budget: 200 USD Deadline: 7 days

Hello!

I have read the technical specifications carefully — the task is clear. Python, Gmail API, parsing PDFs and OCR of images, consolidation between several Excel files, and the final report — all of this is within my stack.

The only nuance worth discussing separately is fuzzy matching between sources: the names of contractors in the bank, accountants, and invoices often differ, so it’s important to agree on a similarity threshold right away to avoid false positives and misses. This can be resolved, but it’s better to fix it at the start. For OCR in Hebrew, I see the Claude API in the specifications — this is the right solution, much better than Tesseract for non-Latin languages.

We will discuss payment details and the exact deadline privately after you share examples of Excel structures and invoice samples — this is needed to provide a clear final estimate, rather than a "guesstimate."

I look forward to collaborating!

Budget: 1150 USD Deadline: 14 days

Good day.

I am ready to create a local Python tool for reconciliation: Gmail API/OAuth, uploading PDFs and images from emails, extracting data from PDFs, OCR for photos/invoices, reading bank and card Excel files, matching by amount/date/contractor with fuzzy matching, and a final Excel report with separate sheets and highlighting of problematic rows.

Estimate: 1000 EUR, in the bid I indicate an equivalent of about 1150 USD, as the project form accepts USD. The timeframe is approximately 14 days.

I will confirm the cost and timeframe after receiving the complete technical specifications and test data: examples of bank statements, cards, accountant files, PDF invoices, and photos of receipts. After that, we can finalize the format of input files, rules for deviations in dates/amounts, categories of recognized expenses, and criteria for report readiness.

I will provide the source code, launch instructions, instructions for Google Cloud credentials, and a test example of the report.

Budget: 100 USD Deadline: 1 day

Good day
I am interested in your task, ready to complete it, feel free to contact me.
I will do it quickly and efficiently, write to me, we will discuss.

Budget: 900 USD Deadline: 7 days

I have a nearly identical project recently completed, but it is for fleet management. I added it as an example - you can take a look.

You can message me privately - I will send more photos, as only one can be uploaded to the portfolio.

It involves working with exports from Gmail, backups to Google Drive, and compiling tables and charts in Google Sheets.

The price depends on the formats and the number of types of PDF files, and how you want to "read" them. If through AI, it’s cheaper but may be less reliable; if extracting with code, it’s more reliable but takes longer to develop and is more expensive.

The price ranges from 350 to 1000 USD, within 7 days.

Budget: 15000 USD Deadline: 7 days

Good afternoon! We have experience in automating financial reporting and parsing bank statements. We implement this through a Python script with a user-friendly interface for matching transactions from different sources. This will eliminate manual errors and speed up the reconciliation process. We are ready to start developing the architecture of the solution immediately after discussing the formats of your data.