Product-led growth has become a mantra in SaaS. Unifyr wants to turn it into measurable profit.

Unifyr announced today the appointment of Holger Schneider as Chief Product Officer, tasking him with aligning product strategy directly to customer outcomes, monetization, and expansion.

The move signals a sharper focus on converting product adoption into sustained revenue growth—an increasingly urgent priority as software buyers demand clearer ROI and faster time-to-value.

From Activation to Expansion

Schneider joins Unifyr with a mandate to strengthen monetization, retention, and expansion across the platform. Working alongside Co-CTOs Lionel Farr and Philip Juchert, he will align Product, Engineering, Design, Data, Marketing, Sales, and Customer Success around a unified operating principle: accelerate activation and help customers realize measurable value sooner.

That cross-functional alignment reflects a broader SaaS trend. Growth is no longer driven solely by acquisition; it’s increasingly tied to adoption depth, usage expansion, and pricing strategy discipline.

Unifyr’s leadership appears intent on tightening the feedback loop between product experience and commercial outcomes.

Strategic Priorities

Schneider’s roadmap centers on five key initiatives:

• Clearer positioning and stronger value propositions across core verticals

• A seamless journey that blends product-led onboarding with sales-assisted growth

• Smarter pricing and packaging aligned with customer success metrics

• Expansion driven by adoption and measurable business impact

• Stronger product marketing, messaging, and competitive differentiation

The emphasis on pricing and packaging is particularly notable. As SaaS markets mature, companies are reevaluating how features translate into value—and how that value is captured.

Usage-based models, tiered packaging, and outcome-oriented pricing are increasingly common. Having a CPO explicitly focused on monetization signals that Unifyr sees product strategy and revenue strategy as inseparable.

The Bigger SaaS Context

The appointment comes at a time when software vendors are under pressure to prove durable growth.

Customer acquisition costs remain high, while investors and boards are scrutinizing net revenue retention and expansion metrics more closely than ever. In that environment, product leaders are being asked to do more than ship features—they’re expected to architect growth engines.

Aligning product, marketing, sales, and customer success around measurable outcomes is easier said than done. Organizational silos often slow value realization and blur accountability.

By formally consolidating that alignment under Schneider’s product leadership, Unifyr is signaling an intent to reduce friction between departments and tighten execution around customer impact.

What It Means for Customers

For customers, the promised outcome is straightforward:

• Faster time-to-value

• Higher product adoption

• Stronger retention

• A scalable growth engine tied to real business impact

If Unifyr can operationalize that vision—connecting product usage directly to customer ROI—it could strengthen both its competitive differentiation and long-term revenue durability.

In today’s SaaS climate, growth isn’t just about adding users. It’s about proving value early, expanding intelligently, and building monetization frameworks that scale with customer success.

With a new CPO focused squarely on those levers, Unifyr is making its priorities clear.

Get in touch with our MarTech Experts.

Klaviyo is putting real capital behind its confidence.

The autonomous B2C CRM provider (NYSE: KVYO) announced that its board has approved a share repurchase program authorizing up to $500 million in buybacks of its Series A Common Stock. As part of that plan, the company intends to enter into a $100 million accelerated share repurchase (ASR) agreement in the near term.

For a growth-stage SaaS company that only recently went public, that’s a notable move.

A Signal of Financial Strength

Share repurchase programs are often associated with mature tech firms generating excess cash. For Klaviyo, the authorization underscores what leadership describes as a “defining year” in 2025—marked by strong revenue growth, expanding profitability, and consistent cash generation.

“This new authorization and accelerated share repurchase underscores the confidence our board of directors and management team have in the durability of our strategy,” said Andrew Bialecki, co-founder and co-CEO.

An accelerated share repurchase structure allows Klaviyo to retire a significant portion of shares upfront, effectively signaling conviction that current valuation levels represent long-term opportunity.

In volatile SaaS markets, that message matters.

Balancing AI Investment and Capital Returns

Klaviyo has increasingly positioned itself as an AI-driven platform, expanding beyond email marketing into a broader autonomous B2C CRM narrative. The company has been investing heavily in AI-powered automation, personalization, and predictive capabilities designed to help brands orchestrate customer journeys across channels.

Bialecki emphasized that the company’s strong balance sheet and cash flow provide flexibility to both invest in AI innovation and return capital to shareholders.

That dual-track strategy—growth plus discipline—has become a defining theme in public SaaS markets. Investors now expect durable margins and capital allocation rigor alongside product expansion.

In recent quarters, many tech firms have faced pressure to curb cash burn and demonstrate profitability. Klaviyo’s buyback announcement positions it closer to the “profitable growth” camp.

Competitive Context in B2C CRM

Klaviyo operates in a crowded customer engagement ecosystem that includes marketing automation platforms, CDPs, and broader CRM suites. The company differentiates itself through deep e-commerce integrations and data-driven personalization tailored to consumer brands.

As AI becomes table stakes in marketing platforms, vendors are racing to embed predictive analytics and automation natively rather than as add-ons.

A strong capital position gives Klaviyo more room to:

• Accelerate AI product development

• Expand platform capabilities

• Pursue strategic partnerships or acquisitions

• Withstand macro volatility

At the same time, repurchasing shares can help offset dilution from stock-based compensation—a common factor in high-growth SaaS companies.

Market Implications

The $500 million authorization does not obligate Klaviyo to repurchase the full amount immediately. Buybacks may occur over time based on market conditions, share price, and capital priorities.

Still, announcing the program—alongside a $100 million ASR—sends a clear message: management believes the stock is undervalued relative to long-term opportunity.

For public SaaS companies navigating shifting investor expectations, capital allocation is increasingly strategic. Buybacks can serve as both financial engineering and psychological reinforcement—signaling maturity and confidence.

The Bigger Picture

The broader SaaS landscape has shifted from “growth at any cost” to sustainable expansion backed by operating leverage. Companies able to demonstrate recurring revenue durability, margin expansion, and disciplined capital management are being rewarded.

Klaviyo’s repurchase plan suggests it sees itself firmly in that category.

The next chapter will hinge on execution—particularly around AI-driven innovation and platform breadth. If Klaviyo can sustain growth while improving profitability, the buyback could look prescient.

For now, the board has made one thing clear: it’s betting on its own strategy.

Get in touch with our MarTech Experts.

Global cloud communications provider Infobip is marking its 20th anniversary with a pivot that signals where it sees the next decade heading: from messaging infrastructure to AI-powered orchestration.

The company this week unveiled AgentOS, a fully managed, AI-native platform designed to operationalize autonomous customer interactions across marketing, sales, and support. Built on top of Infobip’s recently launched AI Agents framework, AgentOS aims to serve as what the company calls a “control layer” for agentic AI—bringing together data, channels, and decision-making into a unified system.

In short: Infobip doesn’t just want to send your messages anymore. It wants to decide, in real time, what should happen next.

From Messaging Pipes to AI Control Plane

Infobip has long competed in the crowded CPaaS (Communications Platform as a Service) market, offering SMS, RCS, email, WhatsApp, and voice capabilities to enterprises worldwide. But as AI agents move from chatbot demos to operational systems, the competitive battleground is shifting.

AgentOS represents Infobip’s evolution from a communications platform to an AI orchestration layer. Rather than building isolated chatbots or campaign automations, enterprises can use AgentOS to deploy goal-driven AI agents capable of managing end-to-end customer journeys autonomously.

That’s a notable leap.

Where traditional marketing automation tools focus on campaigns and prebuilt workflows, AgentOS is built around intent and outcomes. AI agents can determine channel, timing, content, and escalation paths in real time—based on a unified view of customer data.

This shift mirrors broader industry trends. Enterprises are racing to implement generative AI and agentic systems, but many projects stall before reaching production. The culprit is often the same: fragmented data, disconnected systems, and internal governance roadblocks.

Infobip is betting that orchestration—not just intelligence—is the missing link.

Why AI Agent Projects Stall—and How AgentOS Responds

Enterprise enthusiasm for AI agents has been strong, but readiness remains uneven. Many organizations struggle with:

• Unstructured or siloed customer data

• Disconnected marketing, sales, and support stacks

• Security and compliance concerns

• Governance barriers to autonomous automation

AgentOS attempts to address those pain points by combining Infobip’s Conversational Customer Data Platform with real-time journey orchestration. The result is a unified system capable of contextual one- and two-way engagement across more than 15 natively integrated channels.

Instead of stitching together third-party CDPs, CRM systems, and messaging APIs, enterprises get an integrated AI-native platform that connects every touchpoint into a single customer journey.

The promise: fewer disconnected tools, faster deployment, and measurable gains in conversion, satisfaction, and customer lifetime value.

Omnichannel, but Actually Orchestrated

Plenty of vendors claim omnichannel capabilities. What distinguishes AgentOS, according to Infobip, is not simply channel access but orchestration across them.

AI agents can operate autonomously across SMS, RCS, email, WhatsApp, and voice—adapting in real time based on customer behavior and context. If a customer ignores a promotional email, the system might follow up via WhatsApp. If a support inquiry becomes complex, the AI can escalate to a human specialist.

That adaptability hinges on a unified data model and integrated infrastructure. Infobip’s global communications network gives it a foundation that many standalone AI startups lack.

It’s also a defensive play. As hyperscalers and CRM giants embed AI deeper into their ecosystems, communications providers risk being commoditized into message pipes. AgentOS is Infobip’s answer to that threat.

Human-in-the-Loop: Practical AI, Not Sci-Fi

Despite the emphasis on autonomy, Infobip is leaning heavily on a human-in-the-loop model.

AI agents handle scale and repetitive tasks. Human specialists step in for complex cases and continuously refine the models. The approach is pragmatic—particularly in highly regulated industries.

Retail and eCommerce companies are early adopters, using AI agents to deliver hyper-personalized experiences and manage high-volume interactions. But healthcare and financial services firms are moving quickly as well, with a strong emphasis on trust, security, and compliance.

For sectors where a misstep can trigger regulatory scrutiny, fully autonomous AI remains a hard sell. AgentOS’s blended model may make the transition more palatable.

Model Context Protocol: Giving Agents “Communication Superpowers”

One of the more technical but potentially significant elements of AgentOS is Infobip’s integration of Model Context Protocol (MCP) servers.

MCP creates a standardized way for AI agents to interact with third-party systems. In practice, that means agents can move beyond conversation and execute real tasks—booking flights, setting up two-factor authentication, or updating customer records.

Infobip’s MCP servers effectively give AI agents access to its global omnichannel infrastructure. Whether an enterprise uses Infobip’s native agents or third-party models, they can tap into the platform to complete end-to-end, AI-driven customer workflows.

This aligns with the broader shift toward agentic AI: systems that don’t just answer questions, but act.

Modular, API-First, and Built to Scale

AgentOS is built with modular components, open APIs, and MCP interfaces, allowing enterprises to deploy quickly and integrate with existing stacks—or run standalone use cases.

Brands can start with a single application—say, automated customer onboarding—then expand to additional workflows as confidence and ROI grow.

Built-in security and compliance controls aim to address enterprise concerns about data governance and automated decision-making. That’s table stakes in regulated industries, but still a differentiator compared to some AI-first startups that prioritize speed over controls.

Automation and analytics are baked into the platform, feeding continuous optimization loops for personalization and operational efficiency.

Market Context: A Crowded but Evolving Field

Infobip isn’t alone in pushing AI deeper into customer engagement stacks. Major CRM and marketing cloud vendors are rolling out AI copilots and autonomous agents, while CPaaS competitors are layering AI on top of messaging infrastructure.

The question is which layer will own orchestration.

If AI agents become the primary interface between brands and customers, the platform controlling data flow, channel selection, and execution logic could become the most strategic asset in the stack.

Infobip’s global reach and 15-plus natively integrated channels give it scale. AgentOS adds the intelligence layer it previously lacked.

Whether that’s enough to differentiate in an increasingly crowded AI CX market will depend on execution—and enterprise adoption beyond early verticals like retail.

The Bigger Picture: AI as an Operating System for Customer Experience

With AgentOS, Infobip is framing AI not as a feature, but as an operating system for customer experience.

Instead of static workflows and campaign calendars, businesses can deploy goal-driven systems that dynamically respond to customer intent. Instead of siloed departments, they get a unified AI-native layer spanning marketing, sales, and support.

It’s an ambitious shift—from communications vendor to AI control plane.

As agentic AI moves from hype to production, enterprises will need platforms that can manage autonomy without sacrificing compliance or control. Infobip is betting that orchestration—not just intelligence—will define the next era of customer engagement.

And at 20 years old, it’s clear the company isn’t content to remain just a messaging middleman.

Get in touch with our MarTech Experts.

When regulators accept a marketing application, it’s not approval—but it’s a signal. And for Norway-based Photocure ASA and its China-based partner Asieris Pharmaceuticals, that signal just came from Europe.

Asieris announced that the European Medicines Agency (EMA) has accepted its Marketing Authorization Application (MAA) for Cevira (APL-1702), a drug-device combination therapy designed to treat high-grade squamous intraepithelial lesions (HSIL) of the cervix. The milestone clears the application for formal review, moving the product one step closer to potential commercialization in the EU.

For Photocure, listed on the Oslo Stock Exchange under PHO, the development underscores the global expansion potential of a technology platform long associated with photodynamic therapy in oncology. For Asieris, it marks a pivotal regulatory inflection point in Europe.

A Drug-Device Combo Aimed at Avoiding Surgery

Cevira (APL-1702) is positioned as a first-in-class photodynamic therapy for the non-surgical treatment of HSIL, a precancerous condition often linked to persistent high-risk HPV infection. HSIL is typically treated through surgical excision procedures such as LEEP or conization, which, while effective, can carry risks including bleeding, infection, and potential impacts on future fertility.

Cevira aims to change that equation.

As described by Asieris, the therapy combines hexaminolevulinate hydrochloride ointment with a disposable cervical light applicator. The integrated system delivers localized drug administration followed by activation via an intra-cavity cold light source. The result is targeted photodynamic destruction of abnormal cells while preserving surrounding healthy tissue.

In other words, it’s not just a drug. It’s a tightly engineered drug-device system designed to deliver precision therapy directly to the cervix without surgical intervention.

If approved, it could represent a meaningful shift in the standard of care—particularly for women seeking fertility-preserving options.

Backed by Phase III Data With European Representation

The EMA filing is primarily supported by data from an international, multicenter Phase III clinical trial of APL-1702. Notably, more than 20% of enrolled patients were from Europe, a detail that may strengthen the regulatory case by ensuring regionally relevant clinical representation.

While acceptance of the MAA does not imply approval, it confirms that regulators consider the submission sufficiently complete to begin formal evaluation. The EMA will now assess the therapy’s safety, efficacy, and quality data before issuing an opinion.

Given the complexity of drug-device combination products, regulatory review can be more involved than for small-molecule drugs alone. But successful navigation would position Cevira as one of the few non-surgical therapies approved for HSIL in Europe.

Photodynamic Therapy Expands Beyond Bladder Cancer

Photocure has built its reputation around hexaminolevulinate-based photodynamic diagnostics and therapeutics, particularly in bladder cancer. Licensing Cevira to Asieris for development in cervical precancer broadens the platform’s clinical footprint.

The move also reflects a broader industry push toward organ-preserving, minimally invasive oncology treatments. As biopharma increasingly targets earlier-stage disease and precancerous conditions, therapies that reduce surgical burden and improve quality of life are drawing investor and regulatory interest.

For Photocure, the partnership model allows it to leverage its core technology while offloading development and commercialization risk in specific geographies. For Asieris, which is listed on the Shanghai Stock Exchange (688176), the European filing signals ambitions beyond the domestic Chinese market.

Competitive and Market Implications

HSIL represents a significant global health concern, particularly in regions with established cervical cancer screening programs. While HPV vaccination programs are expanding, millions of women continue to require treatment for high-grade lesions detected through screening.

Currently, surgical excision remains the dominant intervention. A non-surgical, office-based therapy that demonstrates comparable efficacy could reshape treatment algorithms—particularly in younger patients.

However, market adoption would hinge on several factors:

• Comparative efficacy versus standard excision procedures

• Recurrence rates over long-term follow-up

• Ease of integration into gynecological practice

• Reimbursement positioning across EU member states

If EMA review concludes positively, Cevira would enter a European market that is increasingly open to device-enabled combination therapies but also highly cost-conscious.

What Comes Next

With the MAA now accepted, the timeline shifts to regulatory review. The EMA’s Committee for Medicinal Products for Human Use (CHMP) will evaluate the application and eventually issue a recommendation. Final approval would then be granted by the European Commission.

For Photocure and Asieris, this phase represents both scrutiny and opportunity.

An approval would validate the clinical program and potentially establish Cevira as the first-in-class photodynamic therapy for HSIL in Europe. A rejection or request for additional data, while not uncommon, could delay commercialization plans.

For now, acceptance of the filing signals that regulators are ready to take a close look at a therapy that aims to reduce reliance on surgery for cervical precancer.

In oncology—and increasingly in women’s health—the future may not just be about treating disease. It may be about treating it earlier, less invasively, and with more precision.

Get in touch with our MarTech Experts.

Asia-Pacific sales tech player Firmable has raised $14 million in Series A funding to take its AI-native sales platform global—starting with a push into the United States.

The round was led by Airtree, with participation from existing investors. The capital will fund US expansion, continued development of Firmable’s proprietary sales dataset, and a deeper build-out of AI agents that don’t just suggest actions—but execute them.

In a crowded sales intelligence market dominated by US-centric datasets and layered SaaS stacks, Firmable’s pitch is pointed: most tools are fast at being wrong, especially outside America.

The Sales Stack Problem

Modern sales teams often operate inside a maze of disconnected tools—CRM systems, enrichment platforms, intent data providers, outreach automation software—each with its own data silo. Coverage is typically strongest in the US and increasingly patchy elsewhere.

The result is a productivity paradox. Reps are armed with more software than ever, yet spend significant time wrangling records, verifying data, and reconciling inconsistencies instead of engaging prospects.

Firmable argues that AI can collapse this patchwork into a single platform—but only if it’s built on proprietary data rather than recycled third-party feeds.

“Most sales intelligence tools are just interfaces on top of the same licensed datasets,” said co-CEO Leigh Jasper. “That’s why the data is stale, duplicated, and US-centric.”

Instead of licensing the same feeds as competitors, Firmable built its own data foundation from the ground up—an expensive and time-consuming strategy, but one that may offer defensibility in a market awash in lookalike AI wrappers.

Three Layers: Data, Signals, Action

Firmable’s platform operates across three tightly integrated layers.

1. Proprietary Data Assembly
The company uses AI-driven web aggregation, large language model-based extraction, and entity resolution to continuously refresh account and contact records. The goal is accurate, structured company intelligence that doesn’t rely on external licensing deals.

2. Precision Buying Signals
On top of that dataset, the platform identifies intent signals such as leadership changes, hiring surges, funding rounds, technology adoption shifts, and other events that may indicate a higher likelihood of purchase.

3. Autonomous AI Agents
The newest focus—and the core of the Series A narrative—is AI agents that act on those signals. These agents can enrich CRM records, prioritize accounts, draft outreach, and orchestrate next steps automatically.

In effect, Firmable wants to move from intelligence to execution.

That positioning aligns with the broader industry shift toward “agentic AI,” where systems perform multi-step tasks autonomously rather than simply generating content or recommendations.

Built in APAC, Expanding to the US

Firmable currently serves more than 1,000 customers across Australia, New Zealand, and eight Asia-Pacific markets. Clients include CBRE, Eftsure, G2, Robert Half, Monday.com, Marsh, and Canon.

The US expansion represents both opportunity and risk. The American sales intelligence market is crowded with incumbents and well-funded startups offering AI-enhanced prospecting tools.

But Firmable sees a strategic angle: most US-built tools struggle internationally. Data coverage thins, workflows assume American go-to-market structures, and accuracy declines.

“Every sales leader we talk to says the same thing: their US-built tools don’t work internationally,” said co-CEO Paul Perrett. “We’re not just filling a coverage gap—we’re building the AI-native platform these teams actually need.”

If Firmable can prove that its data-first model scales into the US while retaining strong APAC coverage, it may appeal to multinational companies frustrated by fragmented global insights.

Founders With a Track Record

Firmable was co-founded by Leigh Jasper, Paul Perrett, and Karthik Venkatasubramanian—veterans of enterprise software company Aconex, which was acquired by Oracle for $1.6 billion.

Jasper, who served as Aconex’s co-founder and CEO, brings credibility in scaling enterprise SaaS from Australia to global markets—a playbook Airtree is betting can repeat.

Airtree partner John Henderson highlighted what he sees as Firmable’s moat: ownership of the underlying dataset. In a market where many AI startups layer models on top of licensed data, proprietary coverage can be a long-term differentiator.

“The AI sales tooling market is exploding, but most startups in the space have no defensible data moat,” Henderson said.

Data as the Moat in AI Sales Tech

The AI sales tooling market has seen rapid growth over the past two years, fueled by generative AI enthusiasm and pressure on revenue teams to do more with less.

But the sector is bifurcating.

On one side are AI copilots layered onto existing CRMs, offering drafting assistance and workflow automation. On the other are intelligence platforms focused on data ownership and signal accuracy.

Firmable is clearly betting on the latter: that AI agents are only as good as the data they’re trained on and act upon.

“Using Firmable, salespeople waste less time on research, eliminate tedious administration, and focus their valuable time on the customers and conversations that matter,” Jasper said.

If that vision holds, the competitive edge won’t be whose LLM drafts the cleverest cold email—it will be who knows which account to target, and when.

The Bigger Picture

As AI matures in sales tech, automation alone is no longer a differentiator. Execution accuracy is.

Firmable’s strategy—own the data layer, surface high-fidelity buying signals, and deploy agents directly on top—positions it as more infrastructure than overlay.

The $14 million raise gives it fuel to test that thesis in the world’s most competitive sales tech market.

If it succeeds, it won’t just prove that AI can collapse the sales stack. It may demonstrate that in the age of autonomous agents, data ownership is the real power play.

Get in touch with our MarTech Experts.

South African operator Cell C is turning to AI to strengthen network continuity—without ramping up infrastructure spending.

The carrier has launched a proof of concept (PoC) with Odine and its wholly owned R&D arm OdineLabs to test an AI-based solution designed to proactively enhance mobile network performance and user experience.

The initiative aligns squarely with Cell C’s capex-light strategy, which prioritizes service quality and customer experience while minimizing heavy infrastructure investments. In a market where spectrum, towers, and equipment upgrades can strain balance sheets, intelligent orchestration is emerging as a more scalable lever.

AI as a Network Quality Multiplier

Rather than focusing on new hardware rollouts, the PoC centers on software intelligence—specifically AI-driven automation and orchestration to improve connection continuity.

OdineLabs’ solution will apply advanced analytics and automation to monitor and optimize network performance in real time. The goal: reduce interruptions, increase reliability, and ensure more consistent connectivity across the mobile network.

For operators, this represents a shift from reactive troubleshooting to predictive optimization. Instead of waiting for customer complaints or performance degradation alerts, AI systems can anticipate congestion, detect anomalies, and adjust network parameters dynamically.

If successful, the PoC could demonstrate how AI can function as a force multiplier—extracting more performance from existing infrastructure.

Supporting a Capex-Light Model

Cell C has publicly emphasized a capital-efficient operating model in recent years, relying on partnerships and shared infrastructure rather than extensive in-house buildouts.

AI-driven orchestration fits neatly into that framework.

By enhancing quality through software layers, the operator can defer or reduce costly physical expansions. That’s particularly relevant in emerging and price-sensitive markets, where revenue per user often doesn’t justify aggressive infrastructure spending.

“Everything we do at Cell C starts with our customers,” said Schalk Visser, CTO at Cell C. “By exploring how AI can proactively enhance network quality, we’re taking meaningful steps toward delivering a more consistent and reliable experience.”

The emphasis on proactive quality management reflects a broader telecom trend: customer experience is increasingly a competitive differentiator, especially as pricing and coverage parity narrow among operators.

Odine’s Africa Expansion Play

For Odine, the partnership strengthens its footprint in Africa and reinforces its positioning as more than a system integrator.

The company’s model blends consultancy, system integration, and AI-powered product development. Through OdineLabs, it aims to bring research-driven innovation directly into commercial telecom environments.

Chairman and CEO Alper Tunga Burak framed AI as foundational to telecom’s next phase.

“We believe AI will fundamentally reshape the way telecom operators deliver value,” he said, adding that OdineLabs is focused on turning next-generation network quality enhancement from a theoretical concept into a scalable reality.

That language echoes a wider industry pivot. As 5G rollouts mature and revenue growth slows, telecom operators are under pressure to improve margins and customer satisfaction without dramatically increasing capital expenditures. AI orchestration, automation, and cloud-native architectures are increasingly central to that strategy.

Cloud-Native, Virtualized, and AI-Orchestrated

The PoC also underscores a shared commitment to cloud-native and virtualized network approaches.

Modern telecom networks are becoming software-defined and increasingly decoupled from proprietary hardware stacks. That shift opens the door for AI engines to sit above the infrastructure layer, orchestrating traffic flows and resource allocation with greater agility.

If Cell C’s PoC validates measurable improvements in continuity and resilience, it could serve as a blueprint for other operators pursuing similar capex-light strategies.

What to Watch

As with any proof of concept, execution and measurable outcomes will determine the next steps. Key questions include:

• How significantly does AI-driven orchestration reduce dropped connections or service interruptions?

• Can the solution scale across broader network segments without operational complexity?

• What cost efficiencies emerge compared to traditional infrastructure upgrades?

For now, the collaboration signals a pragmatic industry reality: telecom’s next performance gains may come less from new towers and more from smarter algorithms.

If AI can meaningfully improve reliability while keeping spending in check, operators like Cell C may find that the most powerful network upgrade isn’t hardware—it’s intelligence.

Get in touch with our MarTech Experts.

At this year’s Mobile World Congress Barcelona, the telecom AI narrative shifted from copilots to collaboration—between machines.

Assurance specialist Mycom announced a strategic partnership with US-based cloud-native networking provider Mavenir to jointly develop Agentic AI use cases for 4G and 5G networks. The goal: move communications service providers (CSPs) beyond dashboard-driven monitoring toward semi- and fully autonomous network operations.

If the industry’s automation rhetoric is to be believed, that’s the promised land.

From Monitoring to Multi-Agent Autonomy

Traditional OSS (Operations Support Systems) platforms have long provided visibility—alarms, performance metrics, fault tickets. What they haven’t reliably delivered is autonomy.

The Mycom–Mavenir partnership is built around Agent-to-Agent (A2A) integration, connecting Mycom’s GenAie NOC Copilot with Mavenir’s Core Domain Intent Agent and domain-specific network function (NF) AI agents inside the mobile core.

Rather than a single AI assistant advising human operators, this model enables multiple specialized agents to collaborate directly—sharing context, diagnosing issues, and triggering remediation workflows across the network stack.

In practice, that could mean:

• Automated detection of network degradation

• Cross-domain root cause analysis

• Closed-loop remediation without manual ticket escalation

The architecture leverages Mycom’s PrOptima (performance management), NetExpert (fault management), and ProAssure (service quality management) platforms alongside Mavenir’s cloud-native mobile core intelligence.

The implication is clear: instead of humans stitching together insights across tools, AI agents do the stitching—and increasingly, the fixing.

Why Agent-to-Agent Matters for 5G

5G networks are inherently more complex than their predecessors. Virtualized cores, distributed architectures, and dynamic slicing introduce operational variables that strain legacy assurance models.

Manual workflows simply don’t scale.

By enabling secure agent-to-agent communication between OSS-level intelligence and domain-native core agents, Mycom and Mavenir aim to create a blueprint for structured, multi-agent collaboration inside live production networks.

According to Mycom Co-founder and CTO Mounir Ladki, the partnership is focused on operationalizing agentic AI “at scale,” not just experimenting with proofs of concept.

The distinction is critical. Many CSPs have tested AI pilots in isolated domains, only to struggle with integration across broader operational ecosystems.

A structured multi-agent framework, if executed cleanly, could reduce the friction between performance management, fault detection, and service quality enforcement.

Accelerating Toward TM Forum Levels 4 and 5

The collaboration also ties directly into the industry’s push toward higher levels of autonomy as defined by the TM Forum Autonomous Networks framework.

Level 4 and Level 5 autonomy envision networks that self-diagnose and self-optimize in real time, with minimal human intervention. Few operators have reached those stages at scale.

Mavenir’s EVP and CTO Bejoy Pankajakshan described the joint initiative as evolving OSS from a monitoring layer into a true autonomy platform—where assurance and optimization occur automatically and continuously.

That framing reflects a broader shift in telecom strategy. As 5G monetization pressures mount and operating margins tighten, CSPs are looking to automation not just for performance gains, but for cost containment.

Autonomous remediation reduces mean time to repair (MTTR), limits service-impacting incidents, and can lower operational expenditure. The business case is as much financial as technical.

Complementary Strengths

The partnership appears strategically aligned.

Mavenir brings deep domain knowledge in the mobile core, along with AI-driven networking solutions embedded directly into cloud-native architectures. Its agents operate close to the network functions themselves, enabling granular insight and closed-loop control.

Mycom, by contrast, sits higher in the OSS stack, offering end-to-end visibility across performance, faults, and service quality.

The agent-to-agent integration effectively connects domain-level intelligence with cross-network orchestration. That layered approach could help CSPs extract more value from existing OSS investments rather than replacing them outright.

In a market where rip-and-replace transformations are both risky and expensive, augmentation through AI may be more palatable.

Industry Context: From Assisted to Autonomous

The telecom sector has spent years pursuing automation through scripts, RPA, and rule-based systems. But rule engines break under the variability of modern, software-defined networks.

Agentic AI introduces a more adaptive model—systems capable of reasoning across context, collaborating with other agents, and taking action based on evolving conditions.

Still, real-world deployment raises questions:

• How are agent decisions audited for compliance and reliability?

• What guardrails prevent cascading automated errors?

• How seamlessly do agents integrate with legacy OSS environments?

These are non-trivial challenges, particularly in live 4G/5G networks supporting millions of subscribers.

If Mycom and Mavenir can demonstrate stable, secure multi-agent operations in production environments, they may provide a credible roadmap for CSPs aiming to reach higher autonomy levels without destabilizing operations.

The Bigger Picture

Telecom operators are under pressure to do more with flat or declining revenue growth. Network complexity is increasing, while tolerance for outages is decreasing.

Agentic AI offers a compelling narrative: networks that detect, diagnose, and fix themselves in near real time.

But autonomy in telecom isn’t a single leap—it’s a series of coordinated integrations across domains.

By formalizing agent-to-agent collaboration between assurance platforms and core network intelligence, Mycom and Mavenir are betting that the path to autonomous operations lies not in one super-agent, but in structured cooperation between many.

If successful, the partnership could mark a meaningful step toward Level 4 and 5 networks—where assurance isn’t just monitored, but executed automatically.

Get in touch with our MarTech Experts.

At this year’s Mobile World Congress Barcelona, three telecom heavyweights laid out a plan to fix one of the industry’s most stubborn pain points: international roaming failures that leave travelers staring at a lifeless signal bar.

NTT DOCOMO, StarHub, and ServiceNow announced a joint initiative to introduce autonomous roaming resolution powered by ServiceNow CRM and the ServiceNow AI Platform. The trio says they are building the industry’s first inter-carrier operational model designed to automatically identify and resolve roaming issues across network boundaries—in real time.

If successful, the project could transform how carriers handle one of the most complex cross-operator workflows in telecom.

The Roaming Problem No One “Owns”

When a customer loses mobile service overseas, the failure rarely sits within a single network. It may involve the home operator, the visited operator, signaling gateways, authentication systems, and clearinghouses. Today, coordination between carriers often relies on fragmented intake channels—web forms, emails, and proprietary portals.

There is no universal, standardized workflow.

For travelers, that can mean hours—or days—without service when they need maps, ride-hailing, or two-factor authentication. For operators, the cost shows up in churn, lost roaming revenue, and brand damage.

In an era where 5G promises ultra-reliability, roaming remains surprisingly manual.

Extending Zero-Touch Beyond a Single Network

DOCOMO has been working with ServiceNow since 2021 to automate internal operations through Zero-Touch Operation (ZTO), eliminating manual intervention in many remote maintenance tasks. That effort reduced fault recovery times and even removed the need for certain overnight support shifts.

Now, the companies are extending that automation model across carrier boundaries.

Instead of handling inter-carrier roaming faults as ad hoc escalations, the new initiative turns them into AI-driven workflows orchestrated on the ServiceNow AI Platform. The system automatically shares structured fault information between participating carriers, tracks resolution progress, and provides real-time visibility into root cause analysis.

In effect, it treats multi-operator troubleshooting as a unified operational domain rather than a patchwork of bilateral agreements.

How the AI ‘Control Tower’ Works

At the core of the initiative is ServiceNow’s AI Platform, acting as a control tower for roaming fault resolution.

When a roaming issue occurs, the system:

• Identifies which network domain is affected

• Pinpoints where the issue originated

• Surfaces relevant performance and fault data

• Automatically routes and tracks resolution tasks across carriers

Instead of multiple human teams exchanging emails and ticket IDs, AI-driven workflows coordinate the process. Fault tickets flow automatically, and recovery actions can begin in near real time.

The approach also promises proactive detection. By analyzing cross-network data, carriers may be able to spot systemic roaming issues before customers begin flooding support lines.

That shift—from reactive to predictive operations—is central to telecom’s broader automation push.

Built on Industry Standards

The collaboration also leans on standards-based interoperability.

The operational model incorporates MEF 113 Trouble Ticketing Business Requirements and Use Cases from Mplify. By grounding the solution in open specifications, the companies aim to reduce fragmentation and make the framework scalable across additional operators globally.

Standards matter in roaming. Without common definitions and processes, automation can’t extend beyond bilateral integrations. If the model proves portable, it could serve as a template for broader industry adoption.

Why This Matters Now

International travel has rebounded sharply, and seamless connectivity is increasingly expected—not appreciated as a bonus.

At the same time, telecom operators face margin pressure and rising operational complexity. 5G cores, virtualized infrastructure, and cross-border traffic flows make troubleshooting more intricate than ever.

Autonomous roaming resolution offers a dual benefit:

• Improved customer experience and trust

• Reduced operational overhead and faster mean time to repair (MTTR)

For operators competing in mature markets, experience differentiation can be as important as pricing or coverage maps.

Technical Validation Underway

The companies confirmed that technical validation is currently in progress, with a commercial launch targeted for the second half of the year.

If deployed successfully, the initiative could signal a broader shift toward cross-carrier automation frameworks—particularly in areas where customer experience depends on coordination beyond a single operator’s domain.

In practical terms, the goal is simple: fewer stranded travelers, more reliable roaming, and standardized inter-carrier processes that scale globally.

In strategic terms, it represents something bigger: a move toward treating telecom operations not as isolated silos, but as interconnected ecosystems managed by AI.

For an industry that has long struggled with fragmentation, that may be the real breakthrough.

Get in touch with our MarTech Experts.