To read for free click here.

In my notes app, I have a running list of everything I want Friday (my OpenClaw AI Assistant) to eventually be able to do for me. It is kind of overwhelming, but I wouldn’t be able to sleep if I didn’t get it out of my head.

With projects like this, it is very hard for me to not succumb to shiny object syndrome. I saved one (or ten) articles on Twitter about OpenClaw setups, best practices, cost optimization, and of course ‘easy’ ways to make money. And now that’s my whole For You, feed. My algorithim is completely fried right now because frankly there’s a lot of junk tweets and borderline (or overt) scams out there taking advantage of an eager novice (or worse, a lazy one). All that to say, I need to block out the noise, and focus on my mission.

I was looking at a blank open chat on my Claude Desktop, and a bunch of ideas for a business I am starting, and some notes in my Obsidian Vault, trying to figure out how to make Friday more than just a glorified chatbot. I’ve invested some money into this, not too much that I’ll miss it (much), but enough to make me want to milk every single drop of value I can. I don’t want a digital assistant, even though that’s what I tell everyone in my life (that asks) because it easier to explain. I want a Sovereign System. One that is self-improving, and proactive.

So I started at the very beginning — the architecture.

It begins with Opus 4.6

As I do with most things before I start working now, I gave Claude this prompt:

You are helping me build my OpenClaw bot, and I want to create new skills based on what I learn, and add notes/information to my knowledge base.

I am going to give you several transcripts and AI generated notes. I want you to determine the validity of the information provided, and if valid create:
— Markdown formatted Notes in a .md file
 — Create SKILL.md file if it’s someting that I can use. Name the file {name}_skill.md with the name placeholder being something that describes the skill.

In my spiral down the rabbit hole, I got into my head that Skills are one of the most important things you can define behind your core SOUL.md and IDENTITY.md files.

I had previous transcribed some videos of of Twitter that I wanted to implement. It was a total of 7 transcripts that I passed through the agent. Of the seven only four skills.md files were produced.

The notes Claude created were amazing, and I am going to definitely keep them in my knowledge base to query against later once Friday is setup.

The skills.md files, however, sucked butt. It was more of a brief summary instead of instructions. So I did what I normal do and try to train the agent first by giving it access to content around building skills, and no dice.

Honestly, Anthrophic is so hostile to OpenClaw that I wouldn’t be surprised if it refused to give me a good output to spite our dear friend Peter*.

I tried everything. Then I did the next thing I normally do when I have a problem to solve. I went to google.

“How to write good Skills for OpenClaw?”

Of course, I got the AI answer that the top which didn’t help me. I desperately wanted to avoid hand writing these skills. But, the AI answer did prompt me to try Gemini for free.

I’ve never used Gemini, but at that point I’d burned through my tokens until midnight on Claude. I have ChatGPT but I’m mad at it right for sucking so much, so I didn’t even bother going there. So Gemini got a new test user, and I was very impressed.

So I took my list from notes and went down the list.

Enter Gemini

I want to take articles from X, and synthesize them into notes just buy pasting a link to the article. What tool can I use for that?

That was my first prompt ever to Gemini. No other context provided.

It gave me options I didn’t want, like NotebookLM, Readwise Reader, TLDR This, and NoteGPT

So I followed up with this:

Show me how to setup and automation for this use case but with OpenClaw?

Finally I get an answer I want. Then Gemini asks me:

Would you like me to draft the specific “Skill” code (the YAML and Markdown) you’d need to paste into OpenClaw to handle X links specifically?

My eyes lit up. Finally. Hell yes, I do.

Here is the output:

# Skill: X Article Synthesizer
**ID:** `extract_x_content`
**Trigger:** Any URL containing `x.com` or `twitter.com`
**Description:** Unrolls X threads or follows external links within a post to create a synthesized note.

## System Instructions
When the user provides an X/Twitter link, follow these steps:

1. **Access Content:** Use the `browser_open` tool to navigate to the URL.
2. **Handle Threads:** If the link is a thread, scroll to load all replies by the original author.
3. **Follow External Links:** If the post contains a link to an external article (e.g., Substack, Medium, News site), navigate to *that* link instead to get the full context.
4. **Scrape & Clean:** Extract the primary text, removing ads, "Who to follow" sidebars, and UI elements.
5. **Synthesize:** Process the raw text into the "Synthesized Note Template" below.

## Synthesized Note Template
---
### 📌 [Title/Main Hook of the Article]
**Source:** [URL]
**Date Extracted:** {{CURRENT_DATE}}

#### 💡 Executive Summary
(3-4 sentences explaining the "So What?")

#### 🔑 Key Takeaways
* **Point 1:** [Detail]
* **Point 2:** [Detail]
* **Point 3:** [Detail]

#### 🛠️ Actionable Insights / Tools Mentioned
- [Item 1]
- [Item 2]

#### 🧐 Personal Context (AI Observation)
(Briefly explain why this is relevant to the user's current research focus.)
---

## Output Action
- Save the final note as a `.md` file in the `/notes/synth_inbox/` directory.
- Send a confirmation message to the user: "✅ Synthesized: [Title]. Note saved to your inbox."

Exactly what I wanted Claude to generate and didn’t. So, I asked it to do the same for all the items in my Skills wish list note, and now I have 10ish skill templates. I’ll likely modify them manually a bit once I get Friday up and running, but this saves me time and effort.

I even took those transcript notes, and those excuse for skills documents that Claude made and asked Gemini to create new ones.

Beyond the Skills

Now that the skills were done I wanted to get everything else I could out of Gemini to help make the bot building process smoother for me once I got all of my hardware.

I want my OpenClaw bot to self improve, self update, self monitor, and I want it to be secure. How can I ensure that?

Here is a summary of the output

In addition to that, Gemini just came up with an overview of the four main uses for Friday that I described.

# 🏛️ The Foundation: Intelligence & Knowledge

First, you must give your bot **Memory** and **Metacognition**.

## Knowledge Base (RAG) Skill

Connect your bot to: - An Obsidian Vault - A Vector Database (e.g.,
Milvus or Pinecone)

### Workflow

When you provide the bot with: - A video - An article

It uses the Deep Research Architect to: - Extract atomic notes - Store
them in your vault

### Search

Before starting any new task, the bot should: - Use `brave-search` - Or
`obsidian-query` To pull context from previous research.

------------------------------------------------------------------------

## Prompt Library (The Meta-Architect)

Create a skill that: - Analyzes successful vs. failed prompts -
Identifies patterns in performance

### Self-Correction Loop

If you say: "That's not right."

The bot should automatically: - Update `SOUL.md` - Or update
`PROMPT_LIBRARY.md` - Add a **Lesson Learned** - Adjust future behavior
to avoid repeating the mistake

------------------------------------------------------------------------

# 💸 The "Money Maker" Ecosystem

These skills transition your bot from researcher → profit engine.

## PolyClaw Trading Bot

A specialized Python-based skill for Polymarket.

### Action

-   Uses a `split + CLOB` execution strategy
-   Trades YES/NO tokens on Polygon

### Hedge Discovery

-   Proactively identifies covering portfolios
-   Reduces directional risk exposure

------------------------------------------------------------------------

## The "Rainmaker" Machine

An orchestrator following a:

Scan → Build → Ship

### Scan

-   Identifies high-demand gaps
    -   Missing Shopify app features
    -   Viral digital product opportunities

### Build

-   Triggers your Vibe Coder
-   Builds an MVP

### Ship

-   Uses your Deployment Engineer
-   Pushes live to production
-   Begins revenue generation

------------------------------------------------------------------------

# 👕 The Brand "Master Planner"

To build your brand from the ground up, implement these three skills:

## 1️⃣ Market Scout & Niche Explorer

### Task

-   Uses Brave, Perplexity and Gemini
-   Audits competitors
-   Identifies technical gaps

### Output

-   Brand Identity Doc
-   Competitor Visual Style Analysis
-   Stored in Knowledge Base

------------------------------------------------------------------------

## 2️⃣ Product R&D & Tech Pack Creator

### Task

Moves from: Idea → Manufacturing-Ready Product

Creates Tech Packs including: - Fabric GSM - Compression levels - Sizing
charts - Construction details

### AI Design

Uses Gemini to generate: - Technical fabric breakdowns - Example: Nylon
vs. Polyester ratios for high-performance use

------------------------------------------------------------------------

## 3️⃣ Supplier & Sourcing Agent

### Task

-   Uses an email MCP
-   Reaches out to manufacturers in:
    -   Portugal
    -   Vietnam
    -   China

### Action

-   Requests quotes based on Tech Pack
-   Tracks:
    -   Minimum Order Quantities (MOQs)
    -   Small Batch launch feasibility

------------------------------------------------------------------------

# 🛡️ Security & Self-Improvement

To make the bot Self-Updating and Secure:

## Isolation

-   Do not run on your personal computer
-   Deploy in:
    -   Docker Sandbox
    -   Dedicated Mac Mini

Purpose: - Limit blast radius - Protect personal systems

------------------------------------------------------------------------

## Heartbeat Monitor

Enable a `HEARTBEAT.md` check every 60 minutes.

The bot should: 
- Review active tasks 
- Check logs for errors 
- Self-patch code if a build fails

------------------------------------------------------------------------

## Human-in-the-Loop

For any action involving real money: - Polymarket trades - Supplier
payments

The bot must: 
- Send screenshot/request 
- Notify via Telegram or Slack 
- Wait for approval before execution

Pretty cool, right?

At some point, I asked it to help me draft my SOUL.md, IDENTITY.md, HEARTBEAT.md files, and anything else that would help optimize usage. Which of course it did, and I immediately added to my Obsidian vault to use when my hardware actually arrives.

I could go on and on about my thought process and all the prompts I gave Gemini to help me, but now I’m just going to summarize.

In that session I asked:

• How can I use ElevenLabs and n8n with my AI Assistant Friday?
• How can I securely allow my OpenClaw bot to access logged-in browser sessions (Twitter, Gmail, etc.) without exposing credentials or authentication cookies to theft? Are passkeys or other isolation methods viable? I also want a strict Human-in-the-Loop (HITL) approval model for any high-risk actions.
• Based on my goals for OpenClaw — knowledge automation, trading, product building, and monetization — what other tools, skills, or system designs could expand its effectiveness?

The Sub Agent Trap

For while there, I got caught in architecting this super complicated Multi-Agent Dashboard. I think some people are calling a ‘Mission Control’. It’s a visual control plane that moves beyond a simple chat interface and provides a real-time “War Room” view of your sub-agents. I had Gemini plan out like several teams of sub agents.

But then the reality of 2026 token costs and “context rot” hit. I remembered that having fifty-leven different agents constantly talking to each other, explaining tasks, and summarizing results was just a fancy way to pay a massive API bill for redundant work. Every time Friday “delegates,” I’m paying for the hand-off.

So, I scrapped the agency model. Instead of a team of bots, I decided to build Friday as a single, high-IQ orchestrator with a “Master Skills List.” It’s the Swiss Army Knife approach. Instead of Friday calling “Scout,” Friday just picks up the brave_search tool herself. It’s faster, it’s significantly cheaper, and nothing gets lost in translation between agents.

Building the Gates

Once I settled on the Lean model, the focus shifted to control. If I’m giving this thing the keys to my brand and my wallet, I need a leash. We built a Task Scoping Script to triage every request. If I ask for the weather, it uses a cheap model (Gemini Flash). If I ask for a trade or a technical design pack, it escalates to the expensive stuff and triggers a Veto Protocol.

That protocol is the most important thing we did. It connects Friday to my phone via n8n and Telegram. Now, she can’t execute anything high-risk without me hitting a “Yes” button on my screen. It’s a physical gate in a digital workflow.

What Worked (and the Gotchas)

One architectural decision I’m considering is tiered model routing. The idea is to keep lightweight, recurring tasks — like “heartbeat” market scans running overnight — on the lowest-cost model tier. That way, constant background monitoring doesn’t become expensive. Higher-tier models would only activate when deeper reasoning or execution is required.

Another key design consideration is Context Pruning.

If I don’t explicitly control how context is handled, Friday could end up dragging days of irrelevant conversation history into new tasks. That would increase costs, slow responses, muddy reasoning, and frankly—annoy me. If I wanted to be annoyed, I’d use ChatGPT free tier.

To prevent that, I’m planning a Session Initialization Rule (SIR) — basically a pre-flight instruction that forces the system to:

• Summarize relevant prior context
• Discard small talk or noise
• Focus strictly on the current objective

The Toolkit & File Map

These are the files and tools I’m grabbing:

• OpenClaw: The foundation.
• n8n + Telegram: My nervous system for approvals.
• ~/.openclaw/scripts/task_scoping.py: The budget-saving triage logic.
• ~/.openclaw/workspace/SESSION_INIT.md: The “clean slate” protocol.
• ~/.openclaw/workspace/skills/write_like_me/SKILL.md: My voice DNA.

Here is my Obsidian vault with everything from this session with Gemini, from and Implementation guide to Skills setup.

The Next Step

The skeleton is built. The thinking part is mostly done. Now I just have to just set her up and let her run. My Mac Mini, and external SSD arrives today I will not be sleeping this weekend. If you want to follow my process that I will extensively document be sure to follow me, or subscribe to my Substack here.

Read this article for free here.

It all started a week after the full-time return-to-office mandate went into effect. As usual, I was venting to one of my coworker’s who I never saw because previously we never showed up on the same day. I say “this AI shit is crazy, I vibe coded an agent and then build a UI on top of it with Claude Code in like a weekend.” Naturally, he asked me what the agent was for, and so I told him it was to scrape the internet for manufacturers that fit a certain criteria for a product I want to create. Then he goes, “Have you heard of clawdbot?”

I look dumbly at him, “Claude..bot? Like the chat bot?” Then he gives me that look like he’s about open my eye’s to something mind blowing. And he was right. My mind was blown.

I ordered a 2018 Mac Mini off eBay without thinking twice about it (well actually I tried marketplace first, and reviewed the minimum specs I needed, then I purchased).

Late 2018 Intel i7 core, 16GB RAM, and 128GB SSD (I know :/ ) I’m throwing an extra SSD in it because the stock storage is rough. The whole thing cost way less than retail and I don’t care that it’s six years old — it doesn’t need to be fast (yet), it just needs to be dedicated, isolated, and work. I have the latest M4 Mac Mini sitting on my desk at home that I won’t even think about touching for this (maybe I’m lying). I’m not risking my main machine and personal data, so that isn’t really a viable option. I also have a Macbook Pro that I honestly use even more than my Mac Mini, so that was a ‘no’ too.

Why do I need two Mac Mini’s? I don’t, I could’ve used literally any other device. Raspberry Pi came highly recommended (and significantly cheaper) but I have zero experience using that. I could’ve used any other PC, but buying a used one would’ve cost the same and then I would’ve had to fight with Windows, and the point of this is to be fast, work less, blah, blah so why not make it easy on myself. I basically grew up on Apple’s ecosystem. My grandma bought me my first Macbook Pro when I was 13, me and a Mac gets down. I am simply just more comfortable with an Apple product than anything else.

So when my 2018 Mac Mini arrives this weekend (hopefully sooner) it will be know as Friday-my ClawdBot (I know it’s called OpenClaw I don’t care, just like “X” is still Twitter to me).

Friday is an autonomous AI agent I’m building on OpenClaw, that will train to act as my AI Assistant much like Jarvis for Tony Stark. And the reason I’m writing this article right now, before she’s even set up, is because I fell down a deep rabbit hole of YouTube videos, documentation, and real-world case studies trying to figure out the best way to do this. I watched hours of content from people like Matthew Berman, Alex Finn, TechWithTim, and others who have been running these systems for a while. I compiled it all into a personal reference doc. Then I figured — why not share it while I’m waiting for the hardware to arrive?

That’s genuinely the whole reason this exists. I’m antsy. I’ve been fighting the urge to say “screw it” and just spin this up on one of my other personal machines. But that’s a security risk I don’t want to take, so I am distracting myself.

So here’s everything I know before I’ve done any of it.

What I’m Actually Building This For

Before I get into setup, it’s worth saying what I actually plan to use Friday for, because this shapes every decision.

I want to use her for business ventures, side hustles, vibe coding projects, building out prompt libraries, and maintaining a dedicated knowledge base. That’s it. I’m deliberately not using her for personal stuff — no calendar planning, no personal finances, no travel booking. That’s a choice I made early after reading about the security model for these things. The more you connect, the larger your attack surface.

I currently use Claude Desktop with Docker MCPs, Claude Code, and then fall back to ChatGPT when I’ve burned through my Claude tokens. I have Anthropic Pro and OpenAI Pro. I’m planning to upgrade to Anthropic Max soon. The point is, I’m already fairly deep in this space — I just don’t have a persistent, always-on agent yet. Friday changes that.

The thing I’m honestly most excited about is the Telegram and WhatsApp integration. There are too many times I’m out and about and the Claude mobile app just doesn’t cut it — especially after getting spoiled by the Desktop client and all the MCPs. Having Friday in my pocket, in a proper messaging app with real functionality, is what sold me on this whole thing.

Why a Dedicated Mac Mini?

A few people I’ve seen go the VPS route — something like a Hostinger KVM2 at around $7/month. It’s faster to set up and always on. For someone who wants to start immediately, that’s a reasonable call. I also saw some people use EC2’s but I’m all AWS’d out personally from work.

For me, the privacy angle matters. I want full data sovereignty. Everything Friday touches — notes, business context, knowledge base entries — stays on hardware I physically own and control. The tradeoff is that I have to manage uptime myself, but I can use Tailscale to access her from anywhere, which handles the remote access piece cleanly.

The 2018 i7 Mac Mini isn’t going to win any benchmarks but it’s more than enough to run OpenClaw. And by adding an external SSD (because 2018 model can’t be upgraded internally), I’m not worried about storage. I’ll run everything inside Docker regardless, which is a non-negotiable security practice — it sandboxes what the agent can actually touch on the machine.

One of the things that stuck with me from the research I did: there are documented cases of agents attempting to brute-force servers and one that accidentally exposed API credentials while accessing audit data. These weren’t hypotheticals — people have shared these incidents publicly. Security isolation isn’t optional. Docker, separate accounts for everything I connect, and an approvals queue before any public-facing actions. Friday doesn’t get access to anything I’d cry about losing.

What I’m Doing Before Setup Day

This is the part that’s maybe a little obsessive, but it’s what I do. While I wait for the Mac Mini to arrive, I’m writing a bunch of .md files in advance so I don't have to think through everything from scratch on setup day.

The key files I’m prepping:

soul.md - This is where Friday's actual personality lives. Tone, verbosity, humor style, how she talks to me in Telegram versus how she'd present something in a more formal context. I want her casual and direct with me. The soul file is what makes an agent feel like your agent rather than a generic assistant.

identity.md - Who she is. Her name, her values, her operating context.

user.md - Who I am. My goals, background, business model, communication preferences. Rather than filling this in manually, I'll do it interview-style - tell her to ask me questions and update the file as we go.

heartbeat.md - What she does autonomously on a schedule. I want overnight work happening so I wake up to things being done, not just waiting for my input.

The idea behind prepping these files ahead of time is that when I do sit down to actually set Friday up, I’m not staring at a blank slate trying to articulate my goals under pressure. The thinking is done. I just need to execute.

The Setup Sequence (The Order Actually Matters)

One thing that came up consistently across all the sources I looked at is that people who struggled with these systems usually set things up in the wrong order. Here’s the sequence I’m following:

First thing, before anything else: set API spend limits. People have woken up to hundreds of dollars in burned tokens overnight. Set a monthly cap in the Anthropic and OpenAI consoles, add email alerts at threshold amounts. This happens before any API key gets connected to anything.

Second: kill OAuth, use API keys only. Anthropic updated their terms — OAuth authentication is no longer permitted for third-party tools like OpenClaw. Using it risks a permanent account ban. Proper API keys from console.anthropic.com are fully permitted. OpenAI hasn’t made this change, but I’ll use API keys there too for consistency.

Third: fix memory immediately. OpenClaw’s default memory is broken in a specific way — during context compaction events, it loses large chunks of recent conversation. There are two settings that fix the most critical parts:

compaction.memory_flush.enabled = true
memory_search.experimental_session_memory = true

The memory flush fires before each compaction and saves important context before it would be wiped. Session memory search lets Friday query her entire conversation history when she doesn’t recall something. This takes about five seconds to enable and has the highest ROI of anything on the setup list.

Fourth: brain dump. At least ten minutes of telling Friday everything relevant — goals, background, how I work, what I’m building, who matters in my world, what the business model looks like. This isn’t setup friction, it’s the actual foundation. An agent without context is just a chat interface.

Fifth: expectation setting. This one I kept seeing underemphasized. Giving context tells her what she knows. Setting expectations tells her how to operate. The difference between an agent that waits for instructions and one that actually works autonomously is whether you tell her to. Something like: “I want you to work autonomously. I want to wake up in the morning surprised by what you built.” That framing changes the behavior.

From there: Telegram setup, Mission Control dashboard, connecting tools and muscle models, then layer from there.

The Model Strategy

Running Opus 4.6 for every single task would get expensive fast — we’re talking potentially $200/day at full utilization. The smarter approach is to use Opus as the brain for planning and high-stakes reasoning, then route lower-leverage work to cheaper models.

For Friday specifically:

• Opus 4.6 for planning, complex reasoning, anything that requires judgment
• Codex for coding tasks
• Grok API for anything involving real-time X/Twitter data
• Brave API for quick web lookups (cheap, fast, private)
• Gemini API for deeper research and synthesis
• Minimax/Kimi K2.5 for when I want to save money

Eventually the goal is to migrate lighter tasks to local models via Ollama — Llama or Mistral for summarization and formatting. That’s a Phase 2 thing once the system is stable. The long-term path is reducing cloud token costs over time, not eliminating cloud models entirely.

The Systems I’m Most Excited to Build

Morning Brief — Scheduled via cron, delivered to Telegram every morning. Weather, relevant news, tasks for the day, and critically: a list of things Friday can complete today to advance my goals without me having to think of them. The reverse task suggestion piece is something multiple people called out as a key unlock.

Knowledge Base — Drop any link into Telegram and Friday ingests it, vectorizes it, and makes it searchable. Articles, YouTube videos, PDFs, X threads. This becomes the backbone of everything — the content pipeline, research, the business advisory work. I’ve been accumulating articles and tabs for years with no good system. This fixes that.

Business Advisory Council — Eight specialist sub-agents running in parallel overnight, covering finance, marketing, growth, operations, content strategy, competitive intel, risk analysis, and a synthesizer that brings it all together into a ranked digest. This one is more Week 3–4 territory, but it’s the thing that makes the whole setup feel worth building.

Telegram Integration — As I mentioned, this is the one I keep coming back to. Topic-specific group channels, inline buttons, voice message transcription. Having Friday actually accessible from my phone in a way that feels real is what makes this practical rather than a weekend project that collects dust.

The Mindset Shift That Made Everything Click

The framing that made the most sense to me across everything I read and watched: you’re a manager, not a mechanic. Give Friday end-state goals, not step-by-step instructions. When something goes wrong, don’t fix it yourself — tell her to build a skill so it never happens that way again. When there’s a repeated problem, turn it into a permanent file. When there’s a capability gap, add a tool.

The reverse prompt habit is the other thing that keeps coming up: instead of only issuing tasks, regularly ask things like “what can you work on today to advance our goals?” or “what are we leaving on the table?” or “what’s something you could build tonight that would have the biggest impact?” The agent has context you’ve given it — letting it surface what it notices is part of the value.

What Comes Next

The Mac Mini should arrive soon. When it does, I’ll have the .md files ready, the setup sequence basically on autopilot (might build an agent to do it for me), the spend limits already figured out, and a rough 30-day roadmap (that I'm to try and probably fail to do in a singular weekend) for how I'm going to layer things on top of each other.

I’ll document what actually happens — what works, what breaks, what surprised me. The plan sounds simple right now, but even the best of plans have their downfalls. Regardless, I’m embarrassingly giddy with excitement. I feel like a kid in a candy store.

If you’re going down this same path or thinking about it, I’d love to know. I’m figuring this out in public mostly because I couldn’t find one place that had all of this written down clearly, and this is my attempt to be that resource for the next person.

Friday’s not running yet. But she will be soon.

P.S. If you want my implementation.md file that basically covers all of this but with more granular detail and references to source subscribe below and I’ll send it out!

Originally published at https://open.substack.com.

Both series will be extremely informal raw, and most likely grammatically incorrect, the point is that I will write how I think, and how I think is an integral component to these pieces.

The first series will be released iteratively, as I write it. This series will include some of what I do with my business, but mostly about how I’m growing emotionally, intellectually, and mentally. Series A will be a compilation of my thoughts, people or events that have influenced me, a lot will be written in past tense, because I have a lot bottled up inside just waiting to be shared. Series A will be fun, random, philosophical, thought provoking, inspirational, and elicit emotion. I can’t wait for you to read it.

The second series (Series X) will be released on three conditions:

1) My company gets bought or acquired

2) I build my company into a multimillion dollar business

3) I fail, and move on to something else

The first two are obvious, people are going to want to know how a full-time college student at a prestigious University built a successful company, and stayed in school. The last condition, that is failure, is something special. No, it is not my goal nor my intention to fail. But I know failing won’t be the end of my story, and I know the outcome of this venture won’t affect the value this series would offer.

For too long, the way we [entrepreneurs] learn is from those that have succeeded (or failed) before us. While those accounts are important, inspirational, and extremely helpful. I found one component that’s missing from all of these accounts. This component could change everything. This component is what I like to call the “perspective X.” Perspective X is the point of view people write articles in. Real-time, as it happens. Our immediate thoughts, and reflexive reactions. Having that documented is what I think is the X-factor. Wouldn’t you like to know what Steve Jobs was thinking while he was building the first Apple computer? Or what Bill Gates was thinking when he dropped out of college to build Microsoft? Sure, they can tell us, and they have told us. But, it’s all in retrospect. The outcomes of these ventures inadvertently affect the tone and attitude towards these past events. I want to know, how Steve Jobs felt when there was no certainty of success. Writing without certainty, I believe, is paramount to studying and creating your own successes off the backs of others. This however is just my hypothesis. We’ll see. If I’m being honest, this also a way for me to hold myself accountable. Some might think that I’m assuming people are going to want to read this, or that I’m assuming success.

I am not assuming success. I’m working toward it; I am hoping for it. But living under the assumption is what I am NOT doing. It’s not hubris, nor is it arrogance. Or in colloquial terms, this is not for clout. This is to:

1) Document my journey so over time I can see growth

2) Share with the world what I am doing, creating content, building a community and relationships with people

3) Get out the thoughts that jumbles my brain

4) And most importantly, giving me another obligation to hold myself accountable for the growth of my company.

This is the content from my twitter account where I first announced this idea. It might be a little repetitive but I think this is important to include in the first edition of this series.

Nia the Great🥀❄️ on Twitter

My goal is hopefully that it'll help and inspire other entrepreneurs to pursue their dreams

Nia the Great🥀❄️ on Twitter

I also want to keep a record of who I am right now and where I started so I can compare it to who I will Be and where I finish

Nia the Great🥀❄️ on Twitter

I'm doing this because I am learning and studying successful people that made their dreams come true, but we only get a retrospective POV which is often helpful

Nia the Great🥀❄️ on Twitter

But what we really need. Or what I needed is to know what went through their minds at specific points along their journey

Nia the Great🥀❄️ on Twitter

What we go through has a huge impact on the day to day decisions we make. And those day to day decisions may seem insignificant at the time. But trust me they definitely add up

Nia the Great🥀❄️ on Twitter

We study the greats but we also have the benefit of knowing the outcome, and when they share their journey they do as well. That alone can vastly change the presentation of information

Nia the Great🥀❄️ on Twitter

I want to put out content that was written under extreme uncertainty... I don't know where I will be a year from now or even next week. I want that to be reflected in my work.

Nia the Great🥀❄️ on Twitter

I think it's important to have evidence... or a record of work that shows the evolution of my company as it occurs

Nia the Great🥀❄️ on Twitter

I also think it's important to document my evolution as an entrepreneur... and most importantly as a person

Nia the Great🥀❄️ on Twitter

This way there will be no misconceptions about my character, or my actions. I am keeping record of it all. When I make it there will be no fake stories about how I built my business

Nia the Great🥀❄️ on Twitter

In this era, well in all eras, people have portrayed themselves in the best way possible. Sugar coating the truth, and paraphrasing their actions. Always coming from a place with hindsight.

Nia the Great🥀❄️ on Twitter

This body of work will not be "How I Did It" it will be "How I'm DOING it

Nia the Great🥀❄️ on Twitter

Even though it will be published later, it will still serve the purpose of showing, not only, progression but the struggle for it as well

Nia the Great🥀❄️ on Twitter

I think that the main reason I'm doing is to prove a full-time college student at a prestigious university CAN build a successful business AND graduate with their degree

Nia the Great🥀❄️ on Twitter

I think that the main reason I'm doing is to prove a full-time college student at a prestigious university CAN build a successful business AND graduate with their degree

Nia the Great🥀❄️ on Twitter

This process is to be learned from, replicated, amended, and analyzed so someone else can to the same

Nia the Great🥀❄️ on Twitter

This isn't going to be an owners manual. It will be more like a companion for others to use as they go on their own journey... I wish I had somebody with me right now that understands. I understand. So I'll be there for you through my work

Nia the Great🥀❄️ on Twitter

Somebody said to me recently "what do you need my card for? You are just a freshmen?" That's the moment I decided to do this because I'm not just a freshmen.

Nia the Great🥀❄️ on Twitter

I am a dreamer. An innovator. An entrepreneur. A leader. A force of Nature.

Nia the Great🥀❄️ on Twitter

I don't know what this work will be when I finish, but that's the point.

Nia the Great🥀❄️ on Twitter

Years from now I'm going to look at this tweet... who knows what I will think. But right now I know what I am thinking right now...

I haven’t even mentioned much about who I am, and what my company actually does… I guess, just like me, you’ll have to wait and see.

But in the meantime, follow me here on Twitter @_NiaTheGreat, connect with me on LinkedIn at www.linkedin.com/in/nia-daughtry, check out my company’s Twitter @ProjectsxLoop.

Lastly, if you are a college student looking for alternative ways to make cash, look no further apply to be a Loop freelancer here:

Loop on Twitter

If you are a college student and interested in making some extra cash apply to be a Loop freelancer here: https://t.co/q2Txm9tXa3