In your first year, you will make dozens of big calls with little time and even less room for mistakes. What you build matters. But what you protect matters too.
If you are building in AI, robotics, or deep tech, your edge is often invisible. It is in your system design. Your training tricks. Your control loops. Your data flow. Your deployment method. Your sensor setup. Your safety logic. Your model updates. Your “small” choices that took months to get right.
A strong IP plan turns those choices into assets. Not paperwork. Not noise. Real assets that make investors take you more seriously, make competitors think twice, and give your team focus on what is truly yours.
This guide will walk through what to do in the first 12 months, step by step, in plain language. You do not need to be an IP expert. You just need a simple plan you can actually follow.
If you want help building this plan with real patent pros, you can apply anytime here: https://www.tran.vc/apply-now-form/
Month 0: Start protecting value before you “have traction”
What “IP” really means in month zero
In the first days, most teams think IP means a patent. But IP is bigger than that. It is anything your company owns that creates an edge. For a deep tech startup, that edge is often the way your system works, not the way it looks.
Your best value may live inside your workflow, your model pipeline, your robot control stack, or the way you turn messy real-world inputs into stable outputs. If you treat that value like a real asset early, you make better choices all year.
Why waiting usually costs you
Many founders tell themselves they will “do IP later,” after they raise or after they ship. The problem is that later is when life gets noisy. You start selling, hiring, fixing bugs, and putting out fires. The clean invention moments from the early days get lost.
Also, your first version often has the most novel ideas. The second version is often refinement. If you only start thinking about IP after refinement, you may miss the strongest part of the story.
The boundary between “what we do” and “how we do it”
You can market and sell without giving away your method. This is the core skill. You should be able to explain the problem you solve, the outcome you deliver, and why it matters, without sharing the exact steps you use.
This boundary is not about being secretive for the sake of it. It is about choosing what you reveal and when. A small amount of planning lets you demo confidently without handing a competitor a blueprint.
A simple way to spot your first patent targets
Ask yourself one blunt question: if a strong rival had your architecture diagram, what would they copy first? That answer is often the first place to explore patents. Another good signal is the thing that took your team weeks to figure out and now feels “obvious.”
When something becomes obvious to you, it is easy to forget how hard it was. That is why founders miss patent chances. They stop seeing the invention because they live inside it every day.
Where Tran.vc fits at this stage
This is the exact moment when Tran.vc can help most. If you want to build protection before a seed round, you can get guided patent strategy and filings as in-kind support, up to $50,000 in value.
If you want to explore that path, you can apply anytime at: https://www.tran.vc/apply-now-form/
Month 1: Set up clean ownership so your future does not get messy
Why ownership is a funding issue, not a legal detail
If the company does not clearly own what is being built, fundraising becomes harder. Investors do not want to guess who owns your code, your designs, or your core method. They want to see a clean story that will hold up later, especially if the company grows fast or gets acquired.
Ownership problems usually do not show up on day one. They show up when you try to raise, sell, or hire senior people. That is when messy history becomes expensive.
Founder agreements that prevent future pain
In the first month, each founder should sign a clear invention assignment agreement. In plain terms, it says that inventions made for the company belong to the company. This keeps the cap table and the IP story aligned.
It also helps between founders. If roles shift later, you avoid arguments about who “owns” the important parts. The company owns them, and the founders own the company.
Contractors are the most common hidden risk
Many early teams use contractors to move fast. That is fine, but you must lock down ownership before work starts. Without the right paperwork, a contractor can legally own what they create, even if you paid them.
A clean agreement is not about distrust. It is about clarity. People do better work when expectations are clear, and you do not want awkward conversations later.
Open source is normal, but you must track it
Most startups use open source, and that is not a problem by itself. The risk appears when a team uses code without knowing the license rules. Some licenses can force you to share your own code if you ship a product in a certain way.
You do not need a complex system at this stage. You need a simple habit: record what you use, why you use it, and the license type. That small step can save months later.
Where Tran.vc fits at this stage
This is the stage where a practical checklist helps more than long meetings. Tran.vc supports founders with clear steps to clean up ownership and reduce risk, while also keeping momentum high.
If you want this support while building your early IP foundation, apply anytime at: https://www.tran.vc/apply-now-form/
Month 2: Build an “invention habit” inside your team
Why invention must be captured while it is fresh
Invention in early startups is messy. It happens during debugging, field tests, and late-night rebuilds. People find a trick, a new method, or a new structure, and then move on. A month later, nobody remembers what was truly new.
If you capture invention while it happens, you build a clean record. That record becomes the raw material for patent work, and it also becomes proof of progress during fundraising.
The two-week invention check-in that does not slow you down
A simple check-in every two weeks can work well. Keep it short and focused. The goal is not to write a perfect document. The goal is to catch the best ideas before they fade.
A good check-in focuses on what changed. What problem was hard before, and what new method made it solvable? If the answer includes a new sequence of steps, a new system design, or a new way to reach a better tradeoff, you likely found something worth exploring.
What counts as an invention in AI and robotics
Founders sometimes think only “big breakthroughs” count. In reality, many strong patents come from practical engineering improvements. A method that makes inference faster on edge devices is valuable. A control approach that reduces failures in real environments is valuable. A system that lowers data needs while keeping accuracy is valuable.
These inventions matter because they connect directly to business value. They reduce cost, improve reliability, and make the product possible at scale.
How this habit improves your team, not just your filings
When your team gets used to naming the “new part,” they build with more intention. They stop adding features that do not create leverage. They notice what is working and why.
This also helps hiring and onboarding. When new people join, they learn the core insights faster, and your company keeps its shape as it grows.
Where Tran.vc fits at this stage
Tran.vc can help you set up a simple invention capture rhythm that fits a fast-moving team. The goal is to support speed while building defensible assets at the same time.
If you want to build this habit with expert guidance, apply anytime at: https://www.tran.vc/apply-now-form/
Month 3: Decide what you will patent and what you will keep secret
The real choice: public protection vs private advantage
A patent is public. A trade secret is private. Both can be valuable, but they work in different ways. In month three, your job is to decide which parts of your edge belong in which bucket.
If you do not make this choice, you may drift into the worst case. You may reveal key details without protection, or you may keep everything secret even when it can be reverse engineered.
What to patent in the first year
A useful rule is to patent what can be copied once your product is in the world. If a competitor can learn it by watching your demo, buying your product, or testing your system, you should consider patent protection.
For robotics, visible mechanisms and repeatable system methods are common targets. For AI, system-level methods often matter more than the model itself, especially if the model can change over time.
What to keep as a trade secret, and how to keep it real
Trade secrets work best for things that stay hidden. This can include internal data pipelines, training recipes, private labeling rules, private datasets, and internal tooling. But trade secrets only work if you can actually keep them secret.
That means controlling access, limiting what gets shared externally, and being careful about what goes into public docs, talks, and code repos. A “secret” that is spread across ten tools and twenty people without controls is not a secret.
The first-year IP map that keeps you focused
A simple map helps: choose a few areas to patent, a few to keep secret, and a few you do not care about. This keeps your team from wasting time trying to protect everything. It also helps you explain your moat clearly to investors.
The map should be revisited as you learn. Your product will change, and your best edge may shift. The point is to stay deliberate, not to be perfect.
Where Tran.vc fits at this stage
This is where experienced guidance is powerful, because the tradeoffs are not obvious. Tran.vc helps founders choose a path that fits their tech, their market, and their speed, while building an IP base that investors respect.
If you want to work through this with a team that does this every day, apply anytime at: https://www.tran.vc/apply-now-form/
Month 4: File your first provisional patents with intention
Why provisional patents matter in the first year
A provisional patent is not the final step. It is a placeholder that gives you an early date while you keep building. In the first year, this matters because your product is still changing, but your core ideas are already forming.
Filing early does not mean locking yourself into a narrow story. It means claiming the ground you are already standing on, so you can build forward with more confidence.
What a strong provisional actually looks like
A weak provisional is vague, rushed, and thin. It names an idea but does not explain how it works. A strong provisional tells a clear technical story. It explains the system, the steps, the parts, and how they interact.
The goal is not to sound impressive. The goal is to make it hard for someone else to say, “This is obvious,” later. Clear writing beats fancy language every time.
How to choose the right scope for early filings
In month four, you should not try to cover everything. Focus on one or two core methods that define your edge. These are often system-level inventions rather than small features.
If you file too narrow, competitors can design around you. If you file too broad without support, the filing becomes weak. The balance comes from describing multiple versions of the same core idea, showing flexibility without losing focus.
Why this step helps fundraising earlier than you expect
Even provisional filings can change how investors see you. They signal seriousness. They show that you think long-term and understand how value is built in deep tech.
You do not need dozens of filings. One or two strong provisionals tied directly to your product story can be enough to change the tone of a seed conversation.
Where Tran.vc fits at this stage
This is where Tran.vc does a lot of hands-on work. They help founders turn raw technical insight into real filings that stand up later, without slowing the team down.
If you want to file early with confidence instead of guessing, you can apply anytime at: https://www.tran.vc/apply-now-form/
Month 5: Align your IP story with your product roadmap
Why IP and product often drift apart
Many teams treat IP as a side task. The product roadmap lives in one doc. The patent work lives somewhere else. Over time, they stop matching. This creates confusion and weakens both efforts.
In month five, your goal is to bring them back together. Your IP should protect where the product is going, not just where it has been.
Using the roadmap to guide future filings
Look at what you plan to build in the next six to twelve months. Where are the hard parts? Where do you expect real engineering risk? Those areas often contain future inventions.
You do not need to file on ideas that are still vague. But you should start watching them closely. When a future feature starts to solidify into a method, it may be time to capture it.
Avoiding patents on features that may be cut
Early roadmaps change. Some features die. If you patent everything on the roadmap, you may waste time and money. The trick is to focus on methods that support multiple directions, not single features.
For example, a flexible system architecture is often more valuable than a narrow use case. The architecture can survive pivots. The use case might not.
Making IP part of product decisions
When the team knows that core methods are protected, they design with more confidence. They are more willing to invest in hard problems instead of quick hacks.
Over time, this creates a product that feels more solid, not just more complex.
Where Tran.vc fits at this stage
Tran.vc works closely with founders to connect IP strategy to real product plans. The goal is not more filings, but better ones that move with the company.
If you want that alignment early, apply anytime at: https://www.tran.vc/apply-now-form/
Month 6: Prepare your IP story for investor conversations
Why investors ask about IP earlier than founders expect
By month six, many teams start having real investor conversations. Even if you are not raising yet, people will ask about defensibility. They want to know what makes your company hard to copy.
A vague answer hurts you more than no answer. It signals that you have not thought deeply about long-term value.
How to explain IP without sounding legal
Your IP story should be simple. You should be able to explain it without legal terms. Focus on the problem, the method, and the advantage.
For example, instead of saying “we have a patent on X,” explain what X enables. Does it make the system cheaper? Faster? Safer? More scalable? Investors care about outcomes.
Showing progress without oversharing
You can say that you have filed, are filing, or plan to file, without sharing details. You do not need to show documents or diagrams. Confidence comes from clarity, not from revealing secrets.
A calm, clear answer builds trust. It shows that you are deliberate and prepared.
Using IP to strengthen your narrative
A good IP story supports your broader vision. It shows that your approach is not a one-off trick, but a foundation you can build on.
This is especially important for AI and robotics, where investors worry about fast followers. Your IP story helps calm that fear.
Where Tran.vc fits at this stage
Tran.vc helps founders shape an IP narrative that fits investor expectations without turning pitches into legal lectures.
If you want help preparing for these conversations, apply anytime at: https://www.tran.vc/apply-now-form/
Month 7: Turn early customer learning into IP insight
Why real-world use reveals new inventions
Once customers start using your product, new problems appear. Edge cases show up. Failures happen. Workarounds emerge. This is fertile ground for invention.
Many strong patents come from solving real customer pain, not from lab ideas.
Watching how your system breaks and recovers
Pay attention to where your system struggles. The fix you design may become more valuable than the original feature. A method that improves stability, safety, or reliability in the field can be a major asset.
These fixes often feel “small” at first, but they are hard-earned and deeply practical.
Capturing customer-driven methods without exposing them
Be careful not to overshare customer-specific details in filings or talks. Focus on the general method, not the specific customer setup.
This keeps your IP broad while respecting customer trust.
Feeding this insight back into your IP plan
Customer learning should update your IP map. Some early ideas may matter less. New ones may rise in importance. Staying flexible keeps your strategy real.
Where Tran.vc fits at this stage
Tran.vc helps founders translate messy real-world learning into clean, defensible IP without slowing down delivery.
If you want to do this well, apply anytime at: https://www.tran.vc/apply-now-form/