Blackstone just wrote a $1.5 billion check to Anthropic. Not for a better chatbot. Not for a new model. For implementation infrastructure that helps enterprises actually use AI in their daily work. The message is clear: the value isn’t in the model, it’s in the last mile of deployment.

If you run a consulting firm, this should change how you think about AI in your own business. The race to integrate the latest LLM into your workflow is a distraction. The real opportunity is building agents that handle the repetitive, high-cost work your senior people do every week. Proposal writing. Client research. Knowledge synthesis across engagements. That’s where the $80K to $300K in annual leakage lives for firms your size.

Blackstone’s bet confirms what we’ve been seeing with advisory and consulting clients over the past 18 months. The firms that get ROI from AI aren’t the ones running GPT-5 demos in partner meetings. They’re the ones who’ve mapped their workflows, identified the bottlenecks, and deployed agents that do specific jobs end to end. Implementation beats models every time.

Why Consulting Firms Waste Time Chasing Models

Most consulting firms treat AI like a technology upgrade. They ask which LLM is best. They wonder if they should wait for the next version. They run pilots that never leave the sandbox. Meanwhile, their senior consultants are still spending 30 hours per proposal, repeating the same research across clients, and watching their knowledge base grow into an unusable archive.

The problem isn’t the model. It’s the gap between the model and the work. A language model can generate text, but it can’t pull your last three healthcare proposals, extract the pricing logic, and draft a new pitch for a pharma client. That requires implementation. It requires an agent that knows where your files live, understands your naming conventions, and has access to your CRM and your Google Drive. The model is a commodity. The implementation is the asset.

Blackstone’s $1.5 billion bet on Ode, Anthropic’s implementation layer, is a signal that the market has figured this out. Enterprise AI value doesn’t come from prompt engineering or fine-tuning. It comes from connecting models to real systems and giving them the context to do real jobs. For consulting firms, that means building agents that handle the work your people hate doing but can’t afford to skip.

The Three Workflows Leaking $80K to $300K Per Year

Let’s get specific. If you’re running a consulting firm doing $1M to $25M in revenue, three workflows are probably costing you more than you realize. Not because your team is inefficient, but because the work is manual, repetitive, and scales badly as the firm grows.

Proposal and pitch development. Every major opportunity starts with a proposal. Your senior people pull past decks, rewrite case studies, adjust pricing, and package it all into a tailored pitch. The work takes 20 to 40 hours per proposal. If you’re bidding on six to ten opportunities per quarter, that’s 500 hours a year of partner time going into documents. At $200 to $400 per hour, that’s $100K to $200K in opportunity cost. Your win rate might be fine, but your cost-of-sale is brutal.

Client research and synthesis. Every engagement starts with research. Industry trends, competitive landscape, regulatory context. Your team spends two to four weeks at the start of each project gathering information, reading reports, and synthesizing it into a brief. The work is necessary, but it’s repeated across clients. You’re paying for the same insight twice because there’s no system to capture and reuse it. Across a dozen engagements per year, that’s another 300 to 600 hours of duplicated effort.

Knowledge management debt. Every project your firm delivers produces intellectual property. Frameworks, models, data sets, client insights. Almost none of it is reusable because it’s locked in individual decks, emails, and meeting transcripts. When a new consultant joins a project, they start from scratch. When a partner needs a reference, they ask around. The firm is paying to generate the same knowledge over and over because there’s no agent reading the corpus and answering questions across it.

Add it up and you’re looking at $80K to $300K in annual leakage. Not from bad process. From manual work that doesn’t need to be manual anymore.

What AI Implementation Actually Looks Like

Here’s what changes when you stop chasing models and start building agents. You pick one workflow, map it end to end, and deploy an agent that does the job without human handoffs. Let’s walk through three examples we’ve built with consulting clients using Omni Ops.

Proposal Generation Agent. This agent lives in your proposal workflow. When a new opportunity comes in, it pulls past proposals for similar clients, extracts relevant case studies, and drafts a tailored document based on your firm’s style and pricing logic. It doesn’t write the final version, but it gives your partner a 70% draft in 20 minutes instead of starting from a blank page. The partner edits, approves, and sends. Proposal time drops from 30 hours to 8. Cost-of-sale drops by two-thirds.

Research Agent. This agent runs at the start of every engagement. You give it a client name and a set of research questions. It searches industry reports, pulls competitive data, reads regulatory filings, and delivers a one-page brief with sources and summaries. The work that used to take your team two weeks now takes 90 minutes. The agent doesn’t replace your analysts, it handles the grunt work so they can focus on synthesis and client interaction. You can read more about how we structure these agents in our AI insights library.

Knowledge Agent. This agent reads everything your firm produces. Every deck, every doc, every meeting transcript. It indexes the content and answers questions across the corpus. When a consultant needs a reference, they ask the agent. When a partner wants to know if the firm has done work in a specific sector, the agent pulls the relevant projects. The knowledge base becomes an asset instead of a liability. You stop paying for the same insight twice.

These aren’t theoretical. They’re agents we’ve deployed with firms in the $2M to $20M range. The implementation takes weeks, not months. The ROI shows up in the first quarter. You can see the full breakdown of how we scope and build these systems on the AI audit for consulting firms.

Why Blackstone’s Bet Matters for Your Firm

Blackstone didn’t invest $1.5 billion in Anthropic because Claude is better than GPT. They invested because Ode solves the implementation problem. It gives enterprises the infrastructure to connect models to their systems, manage agent workflows, and deploy AI at scale without rebuilding everything from scratch. That’s the bet. Implementation is the value layer.

For consulting firms, the lesson is the same. You don’t need a better model. You need agents that integrate with your actual workflows. That means connecting to your CRM, your Google Drive, your project management tools, and your knowledge base. It means building agents that understand your firm’s context, not just general knowledge. It means focusing on the last mile, where the work actually gets done.

The firms that win with AI over the next three years won’t be the ones running the latest LLM. They’ll be the ones who’ve deployed agents that handle the repetitive, high-cost work their people do every week. Proposal writing. Research synthesis. Knowledge management. The work that doesn’t scale but can’t be skipped. That’s where the value is. That’s where the $80K to $300K in annual leakage lives. And that’s where implementation beats models every time.

If you want a practical starting point, we’ve built a worksheet that walks through the process of identifying your first agent use case and scoping the implementation. You can grab it here: Deploy Your First Business Agent. It’s a 20-minute exercise that gives you a clear view of where to start.

How to Scope Your First Agent in 60 Minutes

The hardest part of AI implementation isn’t the technology. It’s deciding where to start. Most consulting firms try to boil the ocean. They want an AI strategy that covers everything. They run workshops, build roadmaps, and end up with a deck that sits in a folder. Nothing ships. Nothing changes.

We take a different approach. Book a 60-min Omni Audit and we’ll walk through your workflows, identify the highest-cost bottleneck, and scope your first agent. You leave with three outputs: a process map, a cost-of-inaction estimate, and a 90-day implementation plan. No deck. No strategy theatre. Just a clear view of what to build and what it’s worth.

The audit is designed for consulting firms doing $1M to $25M in revenue. We focus on the workflows that leak the most time and money. Proposal generation, client research, knowledge management. The work your senior people do that doesn’t need to be manual anymore. We map it, scope the agent, and give you a plan you can execute with your existing team or with our help. You can see the full structure on the AI audit for consulting firms.

Here’s what the audit covers. First, we map your current workflow. Where does the work start? Who touches it? Where does it slow down? We’re looking for the handoffs, the repeated steps, and the places where senior people are doing work that could be automated. Second, we estimate the cost. How many hours per week? What’s the hourly rate? What’s the annual leakage? Third, we scope the agent. What does it need to do? What systems does it connect to? What does success look like? You leave with a clear picture of your first use case and what it takes to deploy it.

Most firms we work with choose proposal generation as their first agent. The ROI is immediate, the workflow is well-defined, and the impact is visible to the whole firm. But the audit will tell you if a different use case makes more sense for your business. The goal is to ship something in 90 days that saves your team 10 to 20 hours per week. That’s the benchmark. If the agent doesn’t hit that, we didn’t scope it right.

Implementation Is the Only Moat

Blackstone’s $1.5 billion bet on Ode is a bet on infrastructure. The models are commoditizing. GPT, Claude, Gemini—they’ll all be good enough for most enterprise use cases within 18 months. The differentiation will come from implementation. The firms that can deploy agents quickly, integrate them with existing systems, and iterate based on real usage will win. The firms that wait for the perfect model will lose.

For consulting firms, this is good news. You don’t need a data science team. You don’t need to fine-tune models. You need to map your workflows, identify the bottlenecks, and deploy agents that do specific jobs. The technology is ready. The models are good enough. The only question is whether you’re willing to focus on implementation instead of chasing the next LLM release.

We’ve built Omni to make this process as simple as possible. You bring the workflow. We bring the infrastructure. You can explore the full platform at Omni or dive into the specific tools we use for consulting firms in our guides library. The goal is to get your first agent live in 90 days, not 18 months. That’s the only timeline that matters.

The firms that treat AI as an implementation problem, not a model problem, will pull ahead over the next two years. They’ll reduce cost-of-sale, scale their knowledge base, and free up senior people to do the work only they can do. The firms that keep chasing models will stay stuck in pilot purgatory. Blackstone just told you which side of that line to be on. The question is whether you’re listening.

Book my Omni Audit and we’ll scope your first agent in 60 minutes. No deck, no strategy document. Just a clear plan and a cost-of-inaction estimate. Let’s build something that ships.