RAQ
Rarely Asked Questions
Less common questions about infrastructure, data boundaries, and AI memory.
Infrastructure & Independence
- Who actually owns the GPU your AI runs on?
- Achiral uses dedicated infrastructure for core AI workloads. The exact deployment model may vary by customer requirements, but customer memory is scoped to that customer's environment.
- Why don’t you just use AWS like everyone else?
- We want more direct control over isolation, cost structure, and deployment choices than a hyperscaler-only architecture provides. Achiral is designed so customer data can remain scoped to the customer's environment and exportable under the applicable agreement.
- What happens to my data if Achiral gets acquired?
- We built Achiral for long-term independence, and we intend to remain that way. Our infrastructure, architecture, and business are designed to avoid dependence on larger platforms wherever possible. But regardless of what happens to us as a company, your organization’s data remains isolated, exportable, and under your control.
- We have a BAA and a privacy contract with a frontier AI, and we are not concerned about models being trained on our data.
- That may be a reasonable choice for some teams. Our focus is narrower: private organizational memory, customer-scoped context, and deployment choices that reduce dependency on a single external AI provider.
Data & Ownership
- If AI learns from my data, who owns what it learns?
- Customer data and customer-specific memory remain scoped to the customer environment under the applicable agreement. We do not pool customer-specific memory across customers or use it to train shared models.
- Do the big AI providers train on my conversations?
- That depends on the provider, product tier, and contract terms. Teams should read their data-use terms carefully, especially when AI providers, infrastructure providers, and model partners are all involved.
- What about my privacy?
- Privacy needs to be designed into the product and the operating model. Achiral focuses on organization-scoped memory, access controls, and isolated environments so teams can use AI with clearer boundaries.
The AI Market
- Why are there so many AI startups if the underlying technology is the same?
- It's a new frontier with lots of opportunities for innovation and differentiation. Companies are experimenting with different approaches to solve multiple problems.
- Is open-source AI actually free?
- The weights may be free. Running them is not. Large models require GPU hardware, inference infrastructure, orchestration, and operational expertise to serve reliably at production scale.
- What’s the difference between an AI that remembers and one that doesn’t?
- A stateless assistant answers from the current prompt and any supplied context. A memory-oriented assistant can preserve operational context over time, which can reduce repeated explanation and improve continuity across workflows.
Philosophy & Positions
- Why “Achiral”?
- In chemistry, an achiral molecule has no inherent handedness—no bias, no predetermined orientation. We named the company after that property because we believe an AI system should start neutral and take its shape entirely from the organization it serves. Not from the internet, not from a general corpus, not from other customers’ data. From yours. Read the full story→
- Why not just use ChatGPT or Claude?
- ChatGPT and Claude are strong general-purpose AI tools. Achiral is designed for teams that also need customer-scoped organizational memory, connected work context, and retrieval boundaries.Chiro can use operational context first and may use external model providers according to the configured routing and customer settings.
- Do you think AI will replace people?
- No. AI will automate tasks, compress workflows, and reduce repetitive work—but organizations are still built on human judgment. Taste, trust, accountability, and decision-making remain human responsibilities. We see AI as augmentation, not replacement. Chiro exists to give teams better memory, so people spend less time retrieving information and more time exercising judgment.
- Why should a small business care about AI infrastructure?
- Because infrastructure determines who controls your intelligence. This matters even more for small businesses, which have to be more risk-aware and have less margin for unpredictable bills, vendor lock-in, or losing access to knowledge their teams depend on. Good AI infrastructure keeps your systems fast, private, predictable, and built to last. Most importantly, it ensures your company’s memory remains yours.
Have a question we haven’t answered?
We’re transparent by default. Ask us anything.