A new Echelon of IT implementations and AI-powered services

Rahul Kayala, Eddie Guo and Anand Sainath are productizing IT services with autonomous RL agents.
Few line items in the IT budget are as staggering as enterprise software implementations. A single ServiceNow deployment can stretch into the tens of millions of dollars. Rollouts of SAP or Salesforce often cost 5x more than the software licenses themselves.
The reality is: implementations, maintenance, and administration are complex workflows. Every company handles their processes in labyrinthine systems of record in totally bespoke ways. As a result, Accenture, Infosys, and Wipro have scaled to tens of thousands of employees, throwing legions of consultants at the problem. For decades, the $1.5T trade has been: the consulting firm delivers expertise, amassed over hundreds of similar projects, in exchange for perpetual support contracts and more downstream projects within your company.
But the model has cracks. Projects run over time and budget. Knowledge walks out the door with consultants. Tariffs and labor shortages chip away at the arbitrage. CIOs grumble but tolerate it, because until now there hasn’t been an alternative.
Echelon’s founders — Rahul Kayala, Eddie Guo, and Anand Sainath — are building a new standard for IT service delivery with AI agents. They first crossed paths while leading core parts of the product and engineering teams at Moveworks, one of the first AI companies built on top of ServiceNow, and where our partner Enrique led the Series A. As Moveworks scaled and eventually agreed to be acquired by ServiceNow, the three of them kept thinking about what else would be possible within IT teams as models kept improving. Ultimately, they realized that if AI models were good enough to handle IT tickets, they could eventually handle entire enterprise projects, from scoping to delivery to maintenance, faster and better than manual consultants.
We intersected with the team at the earliest stages, and invited them to incubate Echelon with us at BCV Labs. Over several weeks meeting with prospective customers, CIOs, and large IT service providers, it became clear that IT consulting would dramatically evolve.
Leading edge models are now capable of executing longer tasks autonomously while reinforcement learning playgrounds and curated data help agents improve on esoteric flows and languages (like ServiceNow Glidescript and multi-nested admin panels). Echelon agents learn from enterprise workflows already in production, and continuously automate more of the ServiceNow administration stack without requirement involvement from a human. Implementations, migrations, and adaptations can happen in days instead of months.
Offshore Consulting as Yesterday’s Breakthrough
The history of IT services is a history of labor models. In the 1970s and 80s, global systems integrators staffed armies of onshore consultants to help enterprises adopt mainframes and early ERP systems. By the 1990s, the breakthrough was offshore delivery. India became the new hub of IT outsourcing, and a generation of firms grew to dominance by arbitraging wages and scaling headcount.
That wave worked — until it didn’t. The economics eroded, timelines stagnated and CIOs lost patience. The model optimized for billable hours, not outcomes.
Rahul, Eddie, and Anand are unusually well-suited to build this company. During their time at Moveworks, they built strong relationships within the ServiceNow ecosystem and earned expertise in driving automation on ServiceNow. Rahul met and worked with hundreds of enterprise CIOs as a product leader, while Eddie and Anand delivered novel AI products as the field rapidly matured, from BERT, to custom SFT models, to an ensemble approach.
A $1T Industry Ripe for Change
We believe their timing couldn’t be better. The amount of labor performed by AI models is doubling every seven months. Reinforcement learning and fine-tuning are enabling models to master highly specific workflows, from configuration management to compliance checks. Enterprises are under pressure to cut costs and show ROI quickly, but their services partners are still selling headcount.
Echelon’s model flips the equation. It replaces offshore contractors with fleets of AI agents that deliver outcomes in a fraction of the time. The platform compounds knowledge across hundreds of projects simultaneously, so expertise scales non-linearly. Partners, SIs and consulting companies that partner with Echelon can validate and guide the agent’s work, creating a hybrid model that is faster and more reliable than traditional offshore consulting.
ServiceNow was the natural entry point. Its implementations are costly and labor-intensive, often requiring large consultant teams to configure workflows, integrate systems, and maintain compliance. Yet the work follows a standardized playbook, making it a strong fit for automation. As ServiceNow expands its own AI capabilities, enterprises face even greater demand for reconfiguration and upgrades, creating a delivery bottleneck the partner ecosystem cannot meet.
Echelon has already established early partnerships with leading SIs and ServiceNow services providers, and has agents managing the ServiceNow platforms at multiple large enterprises. Agents are compressing delivery timelines by an order of magnitude, while delivering higher quality outcomes. Over time, the company will expand into SAP, Workday, Salesforce, and others..
A New Model for Consulting
The opportunity ahead is enormous. Enterprises spend nearly $1.5 trillion each year on IT services, much of it locked in outdated delivery models. If Echelon succeeds, timelines collapse from quarters to days, institutional knowledge compounds across thousands of projects, and enterprises finally free themselves from the inefficiencies of labor-driven consulting. Offshore contractors will give way to fleets of agents that improve with every deployment. It is rare to see a shift this fundamental in how enterprise IT gets done, and we are proud to back Rahul, Eddie, and Anand from the very beginning as they build Echelon into the new operating model for IT services.
