Resources
What we learn in production
Written by the people who deploy the agents. Every number with a source.
Latest
Pieces to start with
Why implementation is the new bottleneck of enterprise AI
95% of enterprise GenAI pilots never touch the P&L. The models are not the problem. The integration, the KPI and the maintenance are. The data, and what it means if you are buying.
Mauricio Moreno
Co-Founder & Co-CEO
The numbers behind it
of enterprise GenAI pilots never touch the P&L
Source: MIT NANDA, The GenAI Divide, 2025
of organizations have scaled AI across business units
Source: Gartner, Sep 2026
of agentic AI projects will be cancelled by the end of 2027
Source: Gartner, Jun 2025
$6.40$0.85
cost per resolved tier 1 ticket, USD · Proof Ledger, Sep 2026
Case: a fintech in México cut cost per ticket from $6.40 to $0.85 in 12 weeks
Sarah Mousa
Content and Proof Ledger
What a voice agent costs in 2026 (and why ours costs $0.035 per call)
Eric Segovia
Platform lead
720x
screening throughput · Proof Ledger, Sep 2026
Case: from 1,000 screening calls a month to 1,000 an hour with an AI interviewer
Sarah Mousa
Content and Proof Ledger
Data first: why we assess your data before we build an agent
Alejandro Hynds
Integrations and data
Agents vs. RPA vs. conversational assistants
Octavio
Deployment Strategist
How we sign a KPI Contract
Mauricio Moreno
Co-Founder & Co-CEO
Forward Deployed Engineers: what they are and why everyone is hiring them
A senior engineer embedded in the client's operation who ships production code and owns the outcome. Where the role comes from, what an FDE does in a week and how to tell one from a contractor with a better title.
Eric Segovia
Platform lead
FDE Playbook
The method, in twelve pages
How a pod takes one process to production in 8 to 12 weeks and leaves your team with everything.
pages
chapters
languages
v1 on 21 October 2026
Why implementation is the bottleneck
The pod
The four phases
The KPI Contract
The handover
Checklists
Webinars
AI Leadership Series
Sixty minutes, one production deployment on screen, questions answered on the record.
- Amazon Bedrock
Agents in production with Bedrock AgentCore
What it takes to run enterprise agents in your own account: runtime, permissions, guardrails, evals and the audit trail a regulator can read.
On stage
- Platform lead, DaCodes
- Guest speaker announced before the session
The KPI Contract, explained
How the baseline is measured, how the target is set, what happens at the commitment gate and the pilots we have stopped.
On stage
- Deployment Strategists, DaCodes
Glossary
The vocabulary of enterprise agents
Defined the way we use them in a contract.
5 terms shown
- A senior engineer embedded in the client's operation who writes production code in the client's repo and owns the business outcome, not a deliverable. The role was born at Palantir; in 2026 it is becoming the standard delivery model for enterprise AI.
- The member of the pod who defines the use case, the baseline, the KPI and the business case together with the client, and runs the commitment gate. The FDE ships; the strategist makes sure what ships is the right thing, measured the right way.
- The clause we sign before building: one KPI, a measured baseline, a target and a date. If the pilot does not move the KPI, the client does not pay for the Scale phase and does not sign the retainer. It is signed at the end of Blueprint, once the data can support the number.
- The decision point at the end of the pilot, typically week 12. One meeting, one page: baseline, target, measured result, eval trend and cost to run. Three outcomes are possible: scale, extend for a defined reason, or stop.
- The monthly service that keeps an agent in production after the pilot: weekly evals, model and policy updates, monitoring, the operations of self-hosted models, and a quarterly business review. Three tiers (Essentials, Professional, Enterprise). It is signed only after the commitment gate is passed.
Which process do we automate first?
Two discovery sessions, no cost, and you leave with two or three quick wins whether or not you build with us.
We reply in under one business day.