Dotwork vs Prisma AIRS AI Gateway
Palo Alto Networks' AI gateway, formerly Portkey, for routing, controlling and observing traffic to LLM providers.
Prisma AIRS AI Gateway sits on the request path between applications and LLM providers, routing traffic, enforcing budgets and limits, and recording cost and usage for each call. Dotwork gives finance and leadership a view of what AI costs the enterprise and what it returns, placing AI spend in the same cost model as the rest of technology.
At a glance
LLM observability
Prisma AIRS AI Gateway
Operating model
Dotwork
Usually runs alongside Dotwork
Dotwork connects to it and puts its data in the context of the whole enterprise.
The short version
What each is built for.
Prisma AIRS AI Gateway
Prisma AIRS AI Gateway is the AI gateway from Palo Alto Networks, formerly known as Portkey. Applications send LLM requests through it to many providers, and it handles routing and fallbacks, budget limits and rate limits, and logs cost, tokens and traces for each request. It can be used as a hosted service or self-hosted.
- Routing LLM requests across providers, with fallbacks
- Enforcing budgets and rate limits on AI usage
- Logging cost, tokens and traces for each request
- Running a single, self-hostable control point for LLM traffic
Dotwork
Dotwork is an AI-native enterprise operating platform built on the Enterprise Operating Model: one live, graph-based model that connects strategy, capital, work, technology and AI across the enterprise. For AI cost, it prices AI usage against the provider price in effect and places that spend in the enterprise cost model, connected to the AI assets, owners and initiatives involved.
- Placing AI spend in the same cost model as the rest of technology
- Pricing AI usage against the provider price version in effect
- Connecting AI assets to owners, findings, spend and funded initiatives
- Showing leaders what AI costs the enterprise and what it returns
The core difference
How Dotwork and Prisma AIRS AI Gateway approach the problem.
Prisma AIRS AI Gateway controls and observes AI traffic as it flows to providers. Dotwork is a management model that places AI spend alongside the rest of the enterprise's cost, ownership and strategy.
Context
The gateway's context is the request in flight: which provider and model it goes to, what limits apply and what it costs. Dotwork's context is the AI estate as part of the operating model: which AI assets exist, who owns them, what they cost and which initiatives they serve.
Decisions
The gateway supports platform and engineering decisions about routing, reliability, limits and per-request cost. Dotwork supports finance and leadership decisions about AI spend, return and accountability.
Time
The gateway acts on each request as it happens, applying budgets and limits in the moment. Dotwork looks at AI spend by period, priced against the provider price version in effect for that period.
Change
In the gateway, teams change routing rules, budgets and limits as usage grows. In Dotwork, AI assets and spend are records that connect to owners, findings and the initiatives that funded them.
Architecture and data model
Side by side.
Prisma AIRS AI Gateway is built to control and observe every AI request in one place, which is what platform teams need to run LLM usage at scale. Dotwork is built to see AI spend in context, connected to the rest of technology cost, the owner of each asset and the initiatives it serves.
| Dimension | Prisma AIRS AI Gateway | Dotwork |
|---|---|---|
| Position | A gateway on the request path between applications and LLM providers | A model of the enterprise that AI spend and assets are loaded into |
| Cost | Per-request cost and tokens, rolled up by application and user | AI spend by model and provider, placed in the enterprise cost model |
| Pricing | Cost calculated from provider prices applied to each request | Usage priced against effective-dated provider price versions |
| Runtime control | Provider routing, fallbacks, budget limits and rate limits | Budget alerts on AI spend; no provider routing |
| Connection to the business | Usage grouped by application, user and metadata the customer attaches | AI assets linked to owners, findings, spend and funded initiatives |
The role of AI
Where the AI gets its context.
The gateway helps teams run AI traffic reliably and within limits. Dotwork helps the enterprise see what its AI estate costs, who owns it and what it returns. They answer different questions and work side by side.
Prisma AIRS AI Gateway
Prisma AIRS AI Gateway is infrastructure for AI traffic. It routes requests across providers, applies budgets and limits, and records cost, usage and traces, as part of Palo Alto Networks' wider AI security portfolio.
Dotwork
Dotwork's assistant answers questions across the whole operating model, and its agents can create and update Dotwork records with optional human review. An MCP server, enabled per site, gives outside AI tools access to the same graph, including AI spend and ownership.
A fair comparison
Where Prisma AIRS AI Gateway is strong — and where Dotwork is different.
Where Prisma AIRS AI Gateway is strong
- It routes requests across many LLM providers, with fallbacks when a provider fails.
- It enforces budget limits and rate limits on the request path, before spend happens.
- It records cost, tokens and traces for each request, by application and user.
- It can be self-hosted, keeping AI traffic and logs inside the customer's environment.
- It sits within Palo Alto Networks' AI security portfolio.
Where Dotwork is different
- Places AI spend in the enterprise cost model, next to cloud, labor, vendor and application cost.
- Prices AI usage against the provider price version in effect for each period.
- Connects AI assets to owners, security findings, spend and the initiatives that funded them.
- Sets AI spend beside the value recorded for each initiative, for finance and leadership audiences.
Choosing
Which fits your situation.
Choose Prisma AIRS AI Gateway when
- You need to route LLM requests across providers with fallbacks
- You want budgets and rate limits enforced on the request path
- You want a self-hostable gateway for all AI traffic
Prisma AIRS AI Gateway fits platform and engineering teams who need one control point to route, limit and observe LLM traffic across providers.
Choose Dotwork when
- Finance needs AI spend in the same cost model as the rest of technology
- Leaders ask what AI costs each initiative and what it returns
- AI assets need clear owners and links to findings and funded work
Dotwork fits finance, technology and AI leaders who need to see what AI costs the enterprise, who owns it and what it returns, in the same model as the rest of technology.
The bottom line
Choose Dotwork when you want to connect the enterprise across systems and silos — with a lower total cost of ownership.
One operating model across strategy, capital, work, technology and AI, instead of another tool to integrate, reconcile and maintain alongside Prisma AIRS AI Gateway.
FAQ
Questions buyers ask.
Does Dotwork route or limit AI requests?+
No. Dotwork does not sit on the request path and does not route requests between providers. It raises budget alerts on AI spend, while a gateway enforces limits as requests are made.
Can Dotwork and Prisma AIRS AI Gateway be used together?+
Yes. The gateway stays the control point for AI traffic, and its usage and cost data can be loaded into Dotwork, where it connects to owners, findings, initiatives and the rest of technology cost. The integration is scoped with each customer.
Is Prisma AIRS AI Gateway the same as Portkey?+
Prisma AIRS AI Gateway is Palo Alto Networks' AI gateway and was formerly known as Portkey. This comparison covers its gateway, cost and observability features.
Keep comparing
Buyers also compare Dotwork with
Be Decisive.Build Fast.Connect Everything.
Start ConnectingSolutions
Platform
Resources
© 2025-2026 Dotwork, Inc. All rights reserved.