Technology

MoEngage Acquires Aampe: The Shift to Per-User AI Agents in Customer Engagement

Martin HollowayPublished 2month ago4 min readBased on 4 sources
Reading level
MoEngage Acquires Aampe: The Shift to Per-User AI Agents in Customer Engagement

MoEngage has acquired Aampe, a startup built around a distinctive architectural idea: assigning a dedicated AI agent to each individual customer instead of managing customers in groups. The deal is all-cash with undisclosed terms. It follows MoEngage's $100 million funding round led by Goldman Sachs Alternatives and A91 Partners in late 2025, bringing the company's total raise to $280 million by mid-June 2026. TechCrunch reported the acquisition on June 23, 2026.

MoEngage says it will fold Aampe's technology into a new "Agentic Customer Engagement Platform" — moving the company away from rules-based campaign logic toward a model where autonomous agents make personalization calls at the individual level.

How Aampe Works Differently

Conventional customer engagement platforms, including MoEngage's current product, let marketers group users into segments, define journey rules, and send messages when users hit certain conditions or fit ML-scored categories. Aampe inverts that structure: instead of applying one campaign logic to millions of users, it runs a lightweight AI agent for each user, learning that person's behavior patterns, channel preferences, and response timing independently.

At scale, this means a brand with five million active users would run five million concurrent agents. The engineering and infrastructure demands are substantial — which explains why MoEngage is treating this as a multi-year platform overhaul backed by new capital, not a quick feature addition.

The Merlin AI Layer

MoEngage has also launched Merlin AI Custom Agents as part of its broader "AI-Native Platform." Aampe's per-user agent technology feeds directly into this layer. The product lets operators set guardrails, channel priorities, and suppression rules — while the individual agents handle the actual decisions: when to send a push notification, whether to offer a discount, how long to wait before trying to re-engage a user.

This separation between operator-defined policy and agent-executed action mirrors how AI agent tools are being built in other enterprise domains — customer support, code generation, sales outreach. Humans set the constraints; agents operate within them.

Capital, Economics, and Regional Context

The $280 million total raise gives MoEngage resources to handle the engineering integration and infrastructure costs of building agent-scale systems. Per-user agents are computationally expensive; the business case only works if the cost per agent drops faster than customer lifetime value increases. That remains an open question, and the broader industry — from Salesforce to Braze to newer AI-native startups — is working through the same math.

MoEngage is an Indian-founded company with a strong customer base in South and Southeast Asia, where mobile-first usage patterns and high app abandonment rates make per-user personalization a pressing commercial need rather than a luxury feature. That regional context matters when evaluating the acquisition's ambition; the use cases driving this move are different from those shaping product roadmaps at US-based MarTech incumbents.

One important distinction: the term "agentic" is being used loosely across software right now. Whether a per-user ML system that optimizes message timing truly functions as an autonomous agent — with the ability to plan toward goals, use tools, and reason across multiple steps — or is simply a more advanced personalization engine, is something the product will need to demonstrate over time. MoEngage's credibility will depend on showing what Merlin AI Custom Agents do beyond the adaptive send-time and content optimization features already standard in the category.

Regardless of terminology, this is a real architectural choice. Building a per-customer agent layer into a customer engagement platform at this scale has not been done at production scale by the major incumbents. Whether MoEngage can execute it and price it in a way that makes the compute costs work for mid-market customers will determine whether this reshapes the competitive landscape or stays a strong positioning story.