How OpenClaw Helps Me With Research & Assignments

Hey everyone. So I were wondered like when I was doing my research paper, I got very very confused because it was too long. So today I'm going to show you

too long. So today I'm going to show you exactly how I automated my research workflow from raw research PDFs to structured assignment drafts using openflow running on WSL and it's

controlled entirely through telegram bot. So there is no copying, no pasting,

bot. So there is no copying, no pasting, nothing just uploading your text and it will do automatically. So let's see how is it possible. So now I will open my command prompt and this is the

powershell. So now as you can see I have

powershell. So now as you can see I have run the open clock gateway status and it's basically giving me the link through which we can actually access the gateway. But before that what I want to

gateway. But before that what I want to actually show you is that for converting any kind of a big research paper you actually need to first uh store into a

dedicated directory. Okay. So what is

dedicated directory. Okay. So what is this dedicated directory here is that you can actually create a dedic dedicated directory for your downloads.

Okay. So this is the downloads by default for Linux. So I can show you my local folder and these are my research papers. 1 2 3 4 5. Okay. So these are

papers. 1 2 3 4 5. Okay. So these are the research papers and then this is the download section. I have created a

download section. I have created a directory for research papers as you can see specifically and then I have copied all the downloads from this uh location

to the research papers directory and then when I check on the history we can see there are five folders of research papers. Now what is the pseudo apt

papers. Now what is the pseudo apt install popular utils? So see we know that opencloud cannot pass the pdfs directly. So we need to first convert

directly. So we need to first convert them to text and if you simply use this command and give your password for Linux it will automatically download it and

then you can convert all the PDFs at once. Okay. So just change your

once. Okay. So just change your directory to research papers and then you can run this command of every file in PDF. You are converting every p every

in PDF. You are converting every p every PDF in the file to text and it's done.

So you'll get a text response here and see this if you want to see your research paper in text format. So see

here this is my research paper completely in the text format. Fine.

See as you can see this is the complete research paper.

See everything is stored here. Very big

research paper but as we can see how easily it is storing here. See it's

copying everything in this bash script.

So let me override and let me exit. So

fine. So now now here comes the powerful part which is controlling everything to telegram. Okay. So I did not I just ran

telegram. Okay. So I did not I just ran the gateway already once. But now let's show you that uh let's go here. Let's go

to Telegram web and see we can I just uploaded a PDF. Okay. before that uh this is just a welcoming command and then I uploaded my PDF. So what is this

PDF? I'll show you. See this is one of

PDF? I'll show you. See this is one of the PDF. See it is almost 13 pages. See

the PDF. See it is almost 13 pages. See

too big, right? So how we can study this much. So that's why using open clock

much. So that's why using open clock does simply summarize the PDF in just one single go right and it's a good thing and also it's not like as we have

already converted it into text. I can

just uh open a random PDF.

Let's see.

So, it's almost 20 pages. So, let me just copy this and then paste it here.

So, I just pasted everything here. Okay.

Now I'll just ask it to create a road map to study this in 1 hour. I have my

hour. I have my research exam in 1 hour and see what it will do.

Let's wait for some time.

So, uh as we can see that uh there is a complete PDF uploaded here.

Okay.

Uh okay. It's a very long PDF to be very frank. Uh but still it will do and see.

frank. Uh but still it will do and see.

So as we can see here, wow, it has basically generated me everything of whatever is the summary here. One by one is generating.

Wow. It's very good thing. So let me just upload a random PDF here again. Uh

so I'll just close this. I'll go to my local folder. You can even like uh keep

local folder. You can even like uh keep it keep a random one and you can also ask it. So uh okay anyway it's in the

ask it. So uh okay anyway it's in the local folder. So I'll just ask it to

local folder. So I'll just ask it to generate me the important questions from this

research paper for uh good research students.

So let's see what it will give.

Let's wait for some time. So how is this happening is that we are actually you know giving the open cloud to read the file and then it sends everything to the

configured LLM and then it replies with the structured extraction. Fine. So I

didn't even ask a structure here. So let

me ask a structure when it gives a response.

So see as we can see we are using a CLI here dedicatedly for open this is a CLI.

So what is the command to install for this? So the command to install a CLI is

this? So the command to install a CLI is a curl command. Okay.

So a curl command is used for here. Uh

it's not here but we can use the curl command to install the CLI. Fine. So

after you install a CLI all these things you can do.

So let's wait for some time.

So see we can that's why we can make all the drafts to save locally inside our WSL Ubuntu right. So yeah, it gave me this response. Fine, it's a good

this response. Fine, it's a good response. Uh I will ask it like uh you

response. Uh I will ask it like uh you know, I'll ask a structured format to be honest. I just ask in a weight way. But

honest. I just ask in a weight way. But

I will just ask it to create a structured assignment on line using the

extracted points.

intro, background, methodology, key findings, discussions,

and at last conclusions.

So, let's see what it gives.

Even I have made a spelling mistake, but I hope it recognizes very well.

And see we can even reuse the prompt again right this has basically a memory context uh location side from that it actually summarizes all your research in

just a few sentences.

So how does workflow automates? So

whenever you have a bulk PDF then you can convert it into text and then you can structure your argument extractions.

You can clear your summaries. You can

make an academic outline generation. You

can use a section wise drafting as well.

Okay. And in the future you can save and see exactly I have given the response and then it is giving me like this.

Okay. Uh why is the formula appearing like this? So see basically it's a text

like this? So see basically it's a text response right? So formula may not be

response right? So formula may not be appearing uh very much uh in the form of uh normal form. It's all in a latex

form. So um I hope currently this issue

form. So um I hope currently this issue gets resolved in the future but I feel it's a very good thing that whenever you want to use you know a provider using

clawbot this is one of the good ways to go unlike chat GPT which gives a very very detailed response it is basically you know giving me the response

according to what I wish so reducing many hours of manual reading into structured output in minutes. See for

example I see this research paper. See

how long it is. If you read this research paper it will take so much time right? It will take so many hours. So

right? It will take so many hours. So

this is a good thing that we are using like open clock to convert everything into just a single page of summary. So

this is how your research paper can get automated and you can study this in a chunk format and you can score good marks in your exams and even you can uh

read your research paper very well. So

that's all from my side. Thank you.