Adobe acquired Rilo for an undisclosed amount, adding natural language workflow automation to its platform following the $1.9 bln acquisition of Semrush. What agency AI architecture

Adobe has acquired Indian startup Rilo, adding natural language marketing automation technology to its platform. The deal includes a team of six and licenses to parts of their technology—financial terms remain undisclosed. The acquisition follows Adobe's $1.9 bln purchase of Semrush and demonstrates how Adobe is building an enterprise AI architecture for marketing: instead of managing individual tasks, brands now get a unified interface for research, analytics, content creation, and distribution.

$1.9 blnAdobe paid for Semrush in 2026
6 peopleRilo team joined Adobe
2025 yearRilo founded
500+ signalsanalyzed by Active Intelligence

Adobe acquires rilo

Rilo enabled users to create workflows in natural language — from competitor analysis and content repositioning to sales call breakdowns and lead generation. The standalone product is being shut down following the acquisition, but the technology will be integrated into Adobe CX Enterprise ecosystem. The company did not specify which exact products will feature Rilo's functionality, but the team will work on AI tools for enterprise marketers.

The logic is straightforward: after acquiring Semrush, Adobe gained access to search visibility data and competitive intelligence, while Rilo adds an automation layer that connects this data to content creation and distribution. Instead of the traditional workflow—analyst exports report → copywriter writes a brief → designer prepares creative → media buyer launches campaign—a brand describes the task in natural language, and AI agents execute the steps sequentially, passing results to each other.

AI agent architecture replaces managing individual tools with managing task chains—a marketer defines the goal, and the system automatically selects the tools and sequence of actions.

What's happening in the martech platform market with AI agents

Rilo isn't the only player automating marketing workflows. ActiveCampaign launched Active Intelligence: Wavelength, which analyzes account history, past campaign results, website activity, and more than 500 business signals to compose emails, build segments, and identify gaps in the customer journey. Archive released Archie—an agent for blogger advertising campaigns that evaluates brand descriptions, analyzes creator content history, verifies audience data, and returns shortlists of vetted candidates.

Auxia introduced Agent Studio — an operating system for marketing processes where agents analyze raw campaign data, map drop-off points in the funnel, generate creative briefs, and automatically push changes to third-party tools. Webflow launched an agency web development platform that converts natural language descriptions into website layouts, writes copy, optimizes code, and maintains design systems. Crescendo released a customer experience management platform where AI models process incoming requests, distribute tasks, and update internal workflows.

How agentic AI systems are transforming influencer and content work

For brands working with influencers, agentic architecture solves the data fragmentation problem. Today, a typical workflow looks like this: an analyst exports reach and engagement metrics from one service, a manager verifies the audience in another, legal reviews the contract in a third, accounting tracks ad labeling in a fourth. AI agents can consolidate these steps: you describe campaign requirements (target audience, budget, topic), the system finds suitable creators on its own, verifies the data, generates a brief, negotiates terms, and oversees publication with proper ad labeling.

Archive Archie is already implementing some of this logic: it analyzes the brand, checks bloggers' content history and audience, and returns a ready-made list. However, full automation requires integration with placement platforms, ad labeling systems, CRM, and financial tools—the kind of ecosystem Adobe is building around CX Enterprise.

Checklist: how to prepare for working with agency AI platforms

  • Structure your historical campaign data: AI agents learn from historical metrics—the cleaner your data (budget, reach, CPM, conversions), the more accurate the predictions.
  • Unify content and creator requirements: create a document with target audience parameters, tone of voice, and prohibited topics—the agent will use it as a baseline filter.
  • Check integrations: make sure your CRM, analytics platforms, and ad labeling systems provide APIs—without them, the agent won't be able to execute actions automatically.
  • Define control points: decide which stages require manual approval (for example, budget sign-off or legal review of the contract).
  • Test on limited campaigns: run the agent on a small segment (for example, finding micro-influencers for a test collaboration) before scaling to your full media plan.

Visibility in Generative Search Engines: A New Metric for Brands

As martech platforms automate workflow, they're embedding tools to track brand visibility in AI-powered search engines. Bazaarvoice launched its AI Visibility Package, formatting product data and user-generated content for generative engines — structuring reviews, product specs, and visual assets so that LLMs can index and cite brands in customer conversational queries.

CrunchJunkie expanded its analytics to monitor brand mentions across generative search platforms — the platform measures visibility, sentiment, and positioning in ChatGPT, Gemini, Perplexity, and Google AI Overviews. Dreamdata released tracking features that show how buyers use conversational platforms in their decision-making journey: it analyzes brand mentions in LLMs, calculates attribution, and assesses how AI recommendations impact conversions. Influencer launched Creator-First Answer Engine Optimization — a platform to track brand recommendations in AI-powered search engines, analyzing how conversational systems cite creator content, measuring social authority, and tracking brand visibility.

On the Russian market, generative search engines haven't become a primary channel yet, but the trend is visible: users increasingly ask Yandex GPT, GigaChat, or Claude instead of traditional search. Brands working with creators should keep this in mind: if a creator's content gets indexed by LLMs, it can influence recommendations in conversational queries. This means that product integration into a review or a mention in a story can drive not just direct reach to followers, but also long-term visibility in AI-generated answers.

Agent system limitations: where manual control is still needed

Agent architecture works well for repetitive processes with clear criteria—finding authors by reach and CPM parameters, auditing audience data, generating standard briefs. But creative tasks (integration concept, video script, non-standard mechanics) and legal matters (negotiating terms with rights holders, ensuring compliance with advertising law) still require human involvement.

The second challenge is data quality. AI agents learn from historical campaigns: if your database lacks structured metrics (for example, all reach figures lumped into one column without separating organic and paid traffic), the system will produce inaccurate forecasts. The third issue is integrations: most blogger management platforms in Russia don't offer open APIs, so an agent can't automatically place an order or verify ad labeling without manual data transfer.

How to measure workflow automation effectiveness

The primary metric is time savings on routine tasks. Compare how many hours your team spent on author research, audience verification, and brief approval before and after implementing the agent. The second metric is error rate: if the agent automatically checks for ad labeling or verifies content against the brief, track the percentage of posts requiring revision.

The third is forecast accuracy. If the system recommends creators based on historical data, track what percentage of recommendations result in successful placements (content was published, metrics matched the forecast, CPM stayed within budget). The fourth is scalability: how many campaigns can your team run simultaneously with an agent compared to manual management.

What Rilo's acquisition means for the Russian market

Adobe CX Enterprise has limited presence in Russia, but the trend toward agentic architecture is relevant for local platforms. Russian influencer agencies and SaaS services for working with bloggers can adopt this logic: instead of a collection of disconnected tools (search, analytics, CRM, ad labeling), offer a single interface where a marketer describes the task and the system executes the entire workflow.

The obstacle is data fragmentation. Reach metrics are stored on one platform, audience data on another, ad labeling information on yet another. Without a unified API standard, agencies will need to integrate with each service separately. The second obstacle is legal specifics: advertising law requires mandatory ad labeling, verification of foreign agent status, and approval from Roskomnadzor for certain categories. An AI agent can automate technical checks (for example, verifying the presence of an OFD token), but final approval remains a lawyer's responsibility.

Frequently Asked Questions

What is agentic AI architecture in marketing

It's a system where a marketer describes a task in natural language, and multiple specialized AI agents sequentially execute steps—from research and analysis to content creation and campaign launch. Unlike traditional tools that require manual switching between platforms, agentic architecture automatically transfers data between stages and selects appropriate tools.

Why did Adobe acquire Rilo after acquiring Semrush

Semrush provided Adobe with search visibility data and competitive analytics, while Rilo adds natural language workflow automation. Together, they enable linking research, content production, and distribution into a single AI-agent-managed workflow—the marketer sets the goal, and the system executes the steps autonomously.

Can blogger selection be fully automated through AI

Partially—AI agents are effective at filtering by quantitative parameters (reach, CPM, audience demographics, content topic). However, final assessment of creative brand fit, verification of legal risks, and negotiation of non-standard terms are still better handled manually. The optimal approach: the agent compiles a shortlist, and the manager selects final candidates.

In brief

  • Adobe acquired Rilo (a six-person team developing natural language workflow automation technology) following its $1.9 billion Semrush purchase — the company is building an AI agent architecture for enterprise marketing.
  • Agent systems replace managing individual tools with managing task chains: a marketer describes a goal, and AI agents sequentially execute research, analysis, content creation, and campaign launch.
  • Martech platforms are adding tools to track brand visibility in generative search engines — CrunchJunkie, Dreamdata, and Influencer measure mentions and recommendations in ChatGPT, Perplexity, and Google AI Overviews.
  • For brands working with bloggers, agent architecture unifies author discovery, audience verification, brief generation, and ad labeling control — but requires structured data on past campaigns and integration with platforms.
  • Key performance metrics for workflow automation: reduction in time spent on routine operations, error rate in publications, forecast accuracy for metrics, and number of campaigns managed simultaneously.
  • Obstacles to adoption in the Russian market include fragmented data across platforms and the need for manual control over legal requirements (ad labeling, foreign agent status, coordination with Roskomnadzor).

CEO comment

Leonid Naumtsev CEO, ETC AGENCY

Adobe's acquisition of Rilo demonstrates how the company is building a concrete architecture for enterprise marketing: instead of a collection of separate tools, a marketer describes a task in natural language, and AI agents execute the entire workflow—from analysis to content creation and distribution. For us, this means we need to urgently structure data from past campaigns and verify API integrations with key platforms, otherwise the agent won't be able to function. What I see as the key trend here is that routine operations—finding creators, validating audiences, generating briefs—can genuinely be automated if the data is clean. However, creative solutions and legal approvals remain our responsibility. The metrics that will help us assess our readiness are: reduced hours spent on routine tasks, error rate in publications, and forecast accuracy for performance metrics. We'll start with a pilot on one campaign type.

ETC AGENCY

ETC helps brands build automated workflows in influencer advertising: from creator discovery and media buying to analytics and KPI forecasting, freeing your team to focus on strategy.

Send a brief