Open Forge: The 'Prove Before You Pay' Model Changing AI Consulting

Open Forge: The "Prove Before You Pay" Model Changing AI Consulting

Most AI consulting engagements begin with a leap of faith. You sign a contract, wire a deposit, and wait months to find out if the solution actually works. If it does not, you are left with a sunk cost and a painful lesson. If it does, you got lucky as much as you got value.

Open Forge was designed to eliminate that leap entirely. It is Felgor's core engagement model, and it works differently from anything else in the AI consulting market: we diagnose your problem, build a working solution, show it to you running in your real context, and only then do you decide whether to move forward.

This article explains exactly what Open Forge is, why it exists, how it works in practice, who it is designed for, and what makes it fundamentally different from a free trial or a proof of concept.

Why the Traditional AI Consulting Model Is Broken

To understand why Open Forge exists, you first need to understand the model it replaces.

The traditional AI consulting engagement follows a predictable sequence: proposal, contract, development, delivery, hope. Each step introduces risk that falls almost entirely on the client.

The Proposal Stage

A consultancy meets with your team, listens to your problem, and goes away for a week or two. They return with a polished proposal that describes what they plan to build, how long it will take, and what it will cost. The proposal is persuasive. It includes architecture diagrams, timelines, and references to similar projects.

But here is the problem: a proposal is a prediction. It is an educated guess about what will work, how long it will take, and what value it will deliver. No matter how experienced the consultancy, they are making assumptions about your data, your systems, your users, and your constraints — assumptions that have not been tested.

The Contract Stage

You sign the contract and wire a deposit — often 30-50% of the total project cost. At this point, the consultancy has been compensated for work that has not been done. They have little financial incentive to deviate from the plan, even if new information suggests they should.

The power dynamic has shifted. Before the contract, the consultancy was working to earn your trust. After the contract, you are hoping they honor it.

The Development Stage

Weeks or months pass. You receive progress updates, screenshots, and status reports. But you are seeing fragments, not a working system. You cannot test it, your team cannot evaluate it, and your stakeholders cannot form an informed opinion about whether this is the right solution.

The Delivery Stage

Finally, the system is delivered. This is the moment of truth — and it is often the moment of disappointment. The solution works, technically. But it does not fit your workflow as well as you expected. The accuracy is lower than promised. The user interface confuses your team. The edge cases were not handled.

You paid in advance for a prediction. The prediction was partially wrong. And now you are in a negotiation about scope, change orders, and additional costs.

The Core Problem

The fundamental flaw in this model is the timing of the commitment. The client commits financially before seeing the solution work in their context. The consultancy commits to a specification before fully understanding the problem. Both parties are locked into assumptions made with incomplete information.

This is not a criticism of any specific consultancy. It is a structural problem with the model itself. Even well-intentioned, highly competent teams produce suboptimal outcomes when the commitment happens before the understanding.

What Open Forge Is

Open Forge is Felgor's answer to this structural problem. It reverses the sequence: we build first, you see it working, then you decide.

The name draws from the metaphor that runs through everything we do at Felgor. A forge is where raw materials are transformed into something precise and useful through heat, skill, and attention. "Open" means the process is visible. You see the work. You evaluate the result. Nothing is hidden behind a contract.

In practical terms, Open Forge means:

• We invest our own time and resources to build a working solution before asking you to invest yours.
• You evaluate a real system, not a proposal, not a prototype, not a mockup — a system running against your real problem.
• Your decision is informed, not speculative. You know what you are paying for because you have already seen it work.
• The risk sits with us, not with you. If what we build does not meet the standard, we absorb the cost of the work already done.

This is not a marketing gimmick. It is how every Felgor engagement operates.

How Open Forge Works in Practice

Open Forge follows four stages: Diagnosis, Build, Demo, and Decision. Each stage has a clear purpose and a clear output.

Stage 1: Diagnosis

Every engagement begins with a deep diagnostic conversation. This is not a sales call. It is an investigation.

Felipe Freitas and Igor Silveira, co-founders of Felgor, are directly involved in every diagnosis. The goal is to understand three things:

1. The real problem. Not the problem as described in an email, but the problem as it exists in your operations. What is the actual pain? What is the measurable impact? What has been tried before?

2. The data landscape. What data do you have? Where does it live? How clean is it? What are the privacy and compliance constraints? Data determines what is technically feasible.

3. The success criteria. What does "working" look like for your team? What metrics matter? What would make this solution worth the investment?

The diagnosis typically involves one to three conversations over the course of a week. At the end, we have a clear picture of whether the problem is one we can solve, and whether the expected value justifies the investment.

What happens if the diagnosis reveals we are not the right fit? We tell you. Not every problem is best solved with custom AI. If an off-the-shelf tool would serve you better, or if the data is not sufficient, or if the expected ROI does not justify the cost, we say so. The diagnosis is free, and honesty at this stage protects both parties.

Stage 2: Build

If the diagnosis confirms a strong fit, we begin building. This is where Open Forge diverges most sharply from the traditional model.

We build on our own time and at our own cost. There is no contract at this stage, no deposit, no financial commitment from you. We are investing in the relationship because we are confident in the quality of what we will deliver.

The build phase focuses on creating a functional version of the solution — not a demo, not a simulation, but a system that works against your actual problem. Depending on the project, this might be:

• A RAG system querying your internal knowledge base and producing accurate, sourced answers
• A classification model trained on your data achieving target accuracy levels
• An automation pipeline processing real inputs and producing real outputs
• A custom AI agent performing a specific workflow end to end

The build phase typically takes two to four weeks, depending on complexity. During this time, we may ask for access to sample data, clarification on business rules, or feedback on early results. The collaboration is active, but the financial risk remains entirely on our side.

Stage 3: Demo

Once the system is functional, we schedule a demonstration. This is not a scripted walkthrough of pre-selected examples. It is a live session where you interact with the system in real conditions.

During the demo, you and your team can:

• Test with your own inputs. Bring the cases, queries, or data that matter most to your business.
• Evaluate accuracy and relevance. Does the system produce correct, useful results?
• Assess the user experience. Is it intuitive? Does it fit your workflow?
• Identify gaps. What edge cases does it miss? What would need to improve?
• Ask technical questions. How does it work under the hood? What are the scaling considerations? What are the ongoing costs?

The demo is the moment of truth. You see the system working — not in theory, not in a proposal, but in practice. This is the information you need to make an informed decision.

Stage 4: Decision

After the demo, you decide. There are three possible outcomes:

1. Move forward. You are satisfied with what you saw, and you want to invest in a production-ready version. We define scope, timeline, and investment together based on the working system — not on assumptions.

2. Request adjustments. The core is strong, but specific areas need improvement. We discuss what changes are needed and what they involve. This is a normal part of the process.

3. Walk away. The solution does not meet your needs or expectations. You do not pay anything. We part on good terms, and the work we invested is our cost, not yours.

There is no pressure. No hard sell. No "limited-time offer." The system either delivers value for your business or it does not. The demo gives you the information to make that judgment clearly.

Why Felgor Takes On the Risk

A reasonable question: why would a consultancy build first and charge later? What is the business logic?

The answer comes down to three factors.

Confidence in Quality

Felipe Freitas and Igor Silveira, co-founders of Felgor, built this model because they trust the work their team produces. When you have genuine confidence in the quality of your output, you do not need the safety net of advance payment. You can let the work speak for itself.

This is not bravado. It is a calculated decision. Felgor is selective about the projects it takes on. The diagnosis stage filters for problems where we have high confidence in our ability to deliver a strong result. We do not accept every inquiry — we accept the ones where Open Forge can work.

Alignment of Incentives

In the traditional model, the consultancy is incentivized to close the deal. Once the contract is signed, the financial outcome is secured regardless of the quality of delivery. This misaligns incentives.

Open Forge aligns them. We only get paid if we deliver something good enough for you to willingly invest in it. Our financial success depends entirely on the quality of our work and the value it creates for you. There is no scenario where we benefit from a mediocre outcome.

Competitive Differentiation

The AI consulting market is crowded and growing more so every year. Many firms offer similar technical capabilities. Open Forge is a structural differentiator that is difficult to replicate. Taking on build risk requires operational discipline, tight project scoping, and genuine confidence — qualities that cannot be imitated by simply changing a pricing page.

An Honest Caveat

Open Forge is not charity. It is a business model. We invest upfront because, historically, the vast majority of clients who go through the full process choose to move forward. The model works because we select the right projects during diagnosis and deliver at a high standard during the build. If we got either of those wrong consistently, the model would not be sustainable.

Who Open Forge Is Designed For

Open Forge is not for everyone. It was designed for a specific profile of client and project. Being clear about this serves both sides.

Ideal Client Profile

Enterprises and established companies. Organizations with real operational complexity, meaningful data assets, and the budget to invest in custom solutions. These are companies where a well-executed AI system can generate substantial, measurable value.

Mature startups. Companies past the early-stage experimentation phase, with product-market fit, stable revenue, and a clear understanding of where AI can create leverage. These are startups that think in terms of ROI, not novelty.

Decision-makers with authority. Open Forge requires direct access to stakeholders who can evaluate the solution and approve investment. Organizations with lengthy procurement processes or committee-driven decisions may find the model less efficient.

Ideal Project Profile

Projects with $20,000+ investment range. Open Forge involves significant upfront investment from our side. The economics work when the project scope justifies that investment. For smaller projects, the overhead of the diagnosis-build-demo cycle is disproportionate.

Problems with clear, measurable impact. The model works best when both parties can objectively evaluate whether the solution delivers value. If the problem is vague or the success criteria are subjective, the demo stage loses its power.

Sufficient data availability. Custom AI requires data. If your organization does not have the data assets to support a custom solution, we will identify that during diagnosis and recommend alternative approaches.

Who It Is Not For

Early-stage startups with limited budgets. If you are pre-revenue or operating on a shoestring, custom AI is likely not the right investment yet. We would rather tell you that honestly than take a project that does not deliver enough value.

Organizations looking for generic AI tools. If your need is for a general-purpose chatbot, a content generation tool, or a standard analytics dashboard, off-the-shelf products are a better fit. Open Forge is for custom solutions to specific business problems.

Projects requiring immediate deployment. The diagnosis-build-demo cycle takes time — typically three to six weeks before you see the working system. If you need something running tomorrow, this is not the right path.

Open Forge vs. "Free Trial": An Important Distinction

Open Forge is frequently confused with a free trial. The two are fundamentally different, and understanding the distinction matters.

What a Free Trial Is

A free trial gives you access to an existing product for a limited time. The product already exists. It was built for a broad market. You test whether it fits your needs. If it does, you subscribe. If it does not, you move on.

Free trials are valuable for evaluating commodity software. But they have nothing to do with custom AI consulting. A free trial lets you test someone else's product. Open Forge lets you evaluate a solution built specifically for your problem.

What Open Forge Is

Open Forge is risk reversal, not a free trial. The distinction matters in several ways:

• Specificity. A free trial gives you a generic product. Open Forge gives you a custom-built system designed for your exact use case.
• Investment of effort. A free trial costs the provider nothing — the product already exists. Open Forge costs us significant time and resources to build something new for each client.
• Scope. A free trial is typically a limited version of a product. Open Forge delivers a functional system that demonstrates the full approach.
• Outcome. A free trial ends with a subscription decision. Open Forge ends with an investment decision based on demonstrated value.

The phrase "prove before you pay" captures the essence better than "free trial." We are proving — through working software — that our approach solves your problem. That is a fundamentally different proposition.

Why the Distinction Matters

Calling Open Forge a "free trial" diminishes what it actually involves. A free trial is low-cost and low-risk for the provider. Open Forge requires us to invest substantial resources into understanding your problem and building a custom solution — with no guarantee of payment. This investment is the reason you can trust the demo: we would not invest our own resources if we were not confident in the outcome.

The Results Open Forge Generates

The Open Forge model produces tangible benefits that go beyond just reducing financial risk. Here is what we observe consistently across engagements.

Faster Decision-Making

When decision-makers can see a working system, they decide faster. There is no need for weeks of internal debate about whether the proposed approach might work. The system is running. The results are visible. The conversation shifts from "should we do this?" to "here is what we are getting — do we want it?"

In practice, the time from first contact to signed engagement is often shorter with Open Forge than with the traditional proposal-contract cycle, despite the additional build step. The demo eliminates the uncertainty that causes procurement delays.

Higher Trust From the Start

Trust in a consulting relationship typically builds over months of working together. Open Forge compresses that timeline. By the time you decide to invest, you have already seen our work, interacted with our team, and evaluated the quality of our output. The relationship starts from a position of demonstrated competence, not promised competence.

This has downstream effects. Projects that begin with high trust experience fewer friction points, faster feedback loops, and better collaboration. The client feels confident providing candid feedback because they have already seen that we respond to it.

Better Project Outcomes

When both parties enter a project with shared understanding of what the system does and how it performs, the production phase is more efficient. There are fewer surprises, fewer scope changes, and fewer misaligned expectations. The working prototype from the build stage becomes the foundation for the production system, not a throwaway.

This means the investment goes further. Instead of spending the first weeks of an engagement aligning on requirements and discovering constraints, those weeks were spent during the Open Forge process. The paid engagement begins with momentum.

More Accurate Scoping and Pricing

Traditional AI projects are notoriously difficult to scope accurately. Estimates made before development begins are frequently wrong — sometimes by a factor of two or three. This leads to budget overruns, scope cuts, or both.

Open Forge provides a natural scoping mechanism. Because we have already built a functional version, we understand the actual complexity of the problem. The production estimate is based on reality, not assumption. This leads to more accurate pricing, which leads to fewer uncomfortable conversations later.

Reduced Buyer's Remorse

Buyer's remorse in consulting is real and common. You sign a large contract, and the next morning you wonder if you made the right call. With Open Forge, this does not happen. You saw the system working. You tested it with your own data. You made an informed decision. There is nothing to regret.

Frequently Asked Questions About Open Forge

What happens if I do not like what is built?

You do not pay anything. That is the core of the model. If the solution does not meet your expectations or does not solve the problem effectively, you walk away with no financial obligation. The investment of time and resources during the build phase is ours to absorb.

How long does the full process take?

The diagnosis typically takes one week. The build phase takes two to four weeks, depending on complexity. The demo and decision can happen within days of the build completing. In total, expect three to six weeks from first conversation to decision point.

What is the typical investment for a project?

Projects start at $20,000 USD. The exact investment depends on the scope and complexity of the production system, which is defined after the demo. Because the demo gives both parties a clear understanding of what the system involves, the pricing conversation is grounded in reality rather than speculation.

Do you build the full production system during Open Forge, or just a prototype?

What we build during Open Forge is a functional system that demonstrates the approach and proves it works in your context. It is not a mockup or a simulation — it processes real inputs and produces real outputs. However, a production system requires additional engineering: security hardening, scalability, monitoring, integration depth, documentation, and support. The production phase builds on the Open Forge foundation and brings it to enterprise-grade readiness.

What kind of data access do you need?

This varies by project. In some cases, we work with anonymized sample data that you provide. In others, we may need access to a representative subset of your actual data through a secure channel. We discuss data requirements during the diagnosis stage and agree on an approach that meets your privacy and security requirements.

Can I share the demo with my team?

Absolutely. In fact, we encourage it. The demo is most valuable when the people who will use the system are in the room. We are happy to present to technical teams, business stakeholders, or executive leadership — whoever needs to evaluate the solution.

What if the project scope changes after the demo?

This happens regularly, and it is a feature of the model, not a bug. The demo often reveals new opportunities or refined priorities that were not visible during the diagnosis. We adjust scope and pricing to reflect the updated understanding. Because both parties are working from a shared reality — the working system — scope discussions are productive rather than adversarial.

Is there any catch?

No. There is no hidden contract, no obligation, no "free but you owe us if..." clause. The model works because we are selective about which projects we take on. We invest in Open Forge engagements where we have high confidence in our ability to deliver value. The projects that convert to paid engagements fund the occasional one that does not.

How to Get Started

If you are considering a custom AI solution for your business and the Open Forge model resonates with how you prefer to evaluate investments, the next step is straightforward.

Book a Free Diagnosis

The diagnosis is the entry point. It costs nothing, commits you to nothing, and gives both parties the information needed to decide whether an Open Forge engagement makes sense.

During the diagnosis, we will:

• Understand your problem in depth — the operational context, the business impact, and the constraints
• Assess feasibility — whether the data, technology, and timeline support a strong solution
• Define success criteria — what "working" looks like in measurable terms
• Recommend an approach — or, if custom AI is not the right path, suggest alternatives honestly

You can book a diagnosis through our website or by writing to [email protected]. Felipe Freitas and Igor Silveira, co-founders of Felgor, are personally involved in every diagnosis. You will not be speaking with a sales team — you will be speaking with the people who will build your solution.

What to Prepare

You do not need a detailed requirements document or a technical specification. Come with:

• A clear description of the business problem you want to solve
• An understanding of what data you have (even a rough idea is fine)
• A sense of the impact — how much time, money, or risk this problem costs you
• Questions — about Open Forge, about Felgor, about custom AI in general

What to Expect After

If the diagnosis confirms a good fit, we will outline the Open Forge engagement: what we plan to build, what data we need, and the expected timeline for the demo. If the diagnosis reveals that Open Forge is not the right path — because the problem is better served by an off-the-shelf tool, the data is insufficient, or the expected ROI does not justify the investment — we will tell you that directly and, where possible, recommend a better alternative.

Either way, the diagnosis gives you clarity. And clarity is valuable whether or not it leads to an engagement with us.

To understand what custom AI is and why it outperforms generic tools, read our complete guide. If you want a concrete action plan, see how to implement AI in 90 days. Book your free diagnosis and see Open Forge in action.

Closing Thoughts

The AI consulting industry has a trust problem. Too many engagements begin with promises and end with disappointment. Too much money changes hands before either party truly understands what the solution will look like.

Open Forge is our attempt to fix that — not with clever marketing, but with a structural change in how the relationship works. We build first. You see it working. Then you decide. The risk sits with us because we believe the quality of our work earns your investment.

This model is not for every consultancy. It requires the discipline to say no to projects where we lack confidence. It requires the willingness to invest our resources with no guarantee of return. And it requires the capability to build something good enough that the vast majority of clients choose to move forward after seeing it.

That is the standard Open Forge holds us to. And it is the standard your business benefits from.

If you are evaluating AI solutions for your business and want to see what is possible before making a financial commitment, the forge is open.