Both Forward Deployed Engineers and consultants work with organizations as external specialists. That's where the similarity ends. The delivery model, accountability structure, output, and pricing are fundamentally different — and choosing the wrong one for an AI deployment project has predictable consequences.
This is the complete comparison.
The Core Distinction
A consultant advises. An FDE delivers.
A consultant's job is to provide recommendations, frameworks, and analysis. They produce a strategy, a roadmap, or a proof-of-concept. They bring expertise to bear on a problem and leave the client with a document or a demo. Execution is the client's responsibility.
An FDE's job is to ship a working production system. They write code, run it, fix it when it breaks, and don't exit until the system is live and the team can own it. Execution is the FDE's responsibility.
This distinction sounds simple. Its implications are extensive.
Accountability Structure
Consultant: Accountable for delivering a work product — a document, a presentation, a proof-of-concept. The work product can be excellent while the project fails. If the company follows the recommendations and the implementation fails, that's outside the consultant's scope.
FDE: Accountable for the outcome — a working production system that does what it was scoped to do. The FDE can't hand off a slide deck and call the engagement complete. If the system isn't in production and working, the engagement isn't done.
This accountability difference changes everything about how the work is done. FDEs make production-grade architectural decisions from day one. Consultants often optimize for defensibility of recommendations rather than for production success.
Integration Depth
Consultant: External. Operates from periodic engagements — workshops, interviews, document reviews, presentations. Has limited visibility into the actual systems, data, and organizational dynamics. Often works from what clients tell them rather than what they observe directly.
FDE: Embedded. Works in the client's Slack, attends their standups, commits to their repository, participates in their on-call rotation. Sees the systems directly, understands the data from working with it, and absorbs the organizational dynamics from operating within them. Context is built through immersion, not interviews.
Timeline and Throughput
Consultant: Timeline is structured around deliverable milestones — "Phase 1: Discovery (4 weeks), Phase 2: Analysis (4 weeks), Phase 3: Recommendations (2 weeks)." The clock runs during the engagement phases. Between phases, the engagement is paused.
FDE: Timeline is structured around production milestones — "Eval framework done (week 4), first integration live (week 7), system in production (week 14)." The clock runs continuously. The FDE is actively building, testing, and debugging throughout the engagement.
Output Comparison
| What you get | Consultant | FDE | |---|---|---| | Production system | No | Yes | | Code (client-owned) | Rarely | Always | | Documentation | Yes | Yes | | Runbooks | No | Yes | | Eval framework | No | Yes | | Knowledge transfer | Extra cost | Built-in | | On-call coverage | No | Yes | | Architecture decisions | Recommendations | Implementation |
Pricing Models
Consultant: Time-and-materials or fixed-fee by phase. Large consulting firms charge $300–$800/hour for senior consultants. Enterprise strategy engagements from McKinsey, BCG, or Deloitte AI practice run $500K–$5M.
FDE: Fixed-scope engagement. You know the total cost before work begins. Typical fdeai.agency ranges:
| Engagement | Duration | Cost Range | |---|---|---| | FDE-Eval | 4–8 weeks | $40K–$120K | | FDE-Agent / FDE-Infrastructure | 8–16 weeks | $120K–$280K | | FDE-Sovereign | 12–24 weeks | $240K–$500K |
The FDE model costs less per outcome than consulting — because the outcome is a production system, not a document.
When to Use Each
Use a consultant when:
- Your organization hasn't decided which AI use cases to pursue
- You need cross-functional executive alignment before any technical work begins
- You need a board-level presentation on AI strategy
- You need independent validation of a technical approach before committing
Use an FDE when:
- You have a defined AI use case and organizational alignment
- You need a working system in production, not a recommendation
- Speed to production is the priority
- You want to build internal capability alongside the system
Warning signs you're using the wrong model:
- You hired a consulting firm to "implement" your AI system and they're producing a roadmap but no code
- You hired an FDE but aren't giving them access to systems, data, and stakeholders
- The engagement is in month 4 and you have no production deployments
The Most Common Mistake
Companies hire strategy consultants when they need engineers. The consulting engagement produces a detailed AI roadmap. The roadmap sits in a slide deck. Nobody implements it because implementation requires dedicated engineering ownership, which the consulting model doesn't provide.
If you have a defined AI use case and organizational buy-in, the fastest path to value is a production deployment engagement — not another strategy layer.
Frequently Asked Questions
Can the same firm provide both consulting and FDE services? Some firms offer both. The important thing is understanding which mode an engagement is in. An "AI implementation" engagement that produces recommendations without working code is consulting, regardless of what it's called.
How do I know if I need strategy consulting or FDE deployment? Ask yourself: "Do I know what I want to build?" If yes, you need an FDE. If no, you may need strategy consulting to define the use case first. Many organizations skip the strategy phase unnecessarily because their use case is clearer than they think.
Are FDE engagements faster than consulting engagements? For reaching a working production system: consistently yes. An FDE engagement takes 8–16 weeks to production. A consulting engagement that produces a recommendation, followed by an implementation phase, typically takes 6–18 months total.
What questions should I ask to distinguish FDE agencies from consulting firms? "Show me a production system you shipped in the last 6 months at comparable scale. What's the uptime? Can I talk to the engineering team who maintained it?"