Budget: 2000 USD Deadline: 3 days
Hello) I am interested in your project. I would be happy to work with you. If anything, feel free to reach out ;)
I am looking to convert old PDF documents from scans to text documents.
You can use OCR to extract and get part of the way, but the document needs to be cleaned up.
I have shown the job as hourly, but in reality it will be by pages, the more pages converted the better, if you are faster then you will make more per hour.
I have different tiers based on the quality of the PDF.
I am including 2 pages as examples of some of the worst cases you will encounter.
Please give examples of your work and your strategy to convert.
Budget: 2000 USD Deadline: 3 days
Hello) I am interested in your project. I would be happy to work with you. If anything, feel free to reach out ;)
Budget: 2000 USD Deadline: 5 days
I am very interested in your project. As a developer who built a custom AI-powered data extraction tool (pdf2table.com), I have a significant technical advantage for this job.
While the public version of my SaaS specializes in exporting to Excel, the core engine runs on advanced LLMs (Large Language Models). This means extracting perfectly clean, plain text from your scans is completely native to my setup. I don't just rely on standard "dumb" OCR; my AI-driven pipeline actually understands the context of the text, allowing it to handle complex layouts and messy scans intelligently.
Because of this, I can process large volumes much faster than a standard data-entry freelancer, with significantly higher accuracy.
Here is my conversion strategy:
1. AI-Powered Extraction: I will run the scanned documents through the backend of my pdf2table engine. The AI will extract the text while intelligently preserving the logical reading order and foundational structure (paragraphs, lists).
2. Contextual Clean-up: Unlike standard OCR, the AI automatically resolves most common artifacts (e.g., fixing broken characters based on context, removing irrelevant page numbers/headers). I will supplement this with targeted scripts for any persistent issues unique to your documents.
3. Manual QA (Quality Assurance): The automated pipeline handles the heavy lifting, giving me the time to manually review and perfect the text, focusing entirely on those "worst-case" degraded pages you mentioned.
Examples of my work:
Since every scanned document is unique, the best proof of my quality is a live test. Please send me the 2 "worst-case" example pages you mentioned in the job description. I will run them through my AI system and manually refine them, sending you the clean text back immediately. This will demonstrate both my speed and the final quality you can expect.
I am ready to handle high volumes and scale the process as needed. Let's discuss the tiers and get started on the test pages!
article listing 8-10K characters Topic: TOP interior design studios with "turnkey" project implementation in Khmelnytskyi Specifications https://docs.google.com/document/d/15Q2i0LtZO5mq6OKM_16tFaMnj8dM_2RtSuxNqIKIh-w/edit?usp=sharing Project - Soprano Requirements: Text uniqueness >80% No use of AI (ChatGPT, ..) copywriting. Resource for text checking https://www.zerogpt.com/?gclid=CjwKCAiAtt2tBhBDEiwALZuhAC71DYp3iGrmizSErwIfTBiuzeVKLbUuLgink_7txkbHKafTWWDyjBoCCCYQAvD_BwE
We are looking for a specialist with experience who will analyze scientific articles, make editorial corrections if necessary, and submit them to journals indexed in Scopus, ensuring acceptance within optimal timeframes. The workload is from 5 to 15 projects per month. Fair payment for each project. Please describe your experience.
I am looking for a service to publish one or several Wiki articles based on a template that's already pre-written + potentially writing new ones, with my approval. You must speak fluent English. I speak Russian, I am sorry not Ukrainian.
It is necessary to write many different SEO texts. Not manually, but through ChatGPT or other neural networks. The point is that it's not just about throwing in keywords and letting it write - no. We can do that ourselves. It is necessary to clarify many times what is wrong in the text. Communicate simultaneously with Gemini and, for example, ChatGPT so that one points out the mistakes of the other. As a result, the texts turn out to be expert, clear, unique (checked by a special website - we will provide it to you), and most importantly, high quality. Message me privately if you are ready to take on this task. We will discuss the details.
There is a website. I will send it in a private message. Content needs to be written and created for it. Then it should be filled on the site. This is a web studio site. The task is to create descriptions for projects, services, the main page, the team, etc. In other words, to fully describe everything.