Home Blog Page 469

Nymbus Launches Industry-Leading, Secure MCP Server for AI-Driven Core Banking Actions

info@uweb3.io
-
0
Nymbus Launches Industry-Leading, Secure MCP Server for AI-Driven Core Banking Actions

WHY THIS MATTERS: The introduction of a purpose-built Model Context Protocol (MCP) for the core is a critical architectural pivot in the evolution of agentic banking. For years, financial institutions have struggled to bridge the chasm between flexible, cloud-native AI tools and their rigid, legacy core banking systems, fearing compliance gaps and operational risk. This server provides a secure, standardized connection, finally making AI-assisted workflows practical for everyday front-office operations. By translating conversational AI requests into approved core functions, Nymbus is laying the groundwork for a new competitive advantage: the ability to execute complex actions like fraud investigation or account management through a single interface, delivering genuine efficiency gains instead of mere chatbot answers. This move signals that responsible AI integration, rather than just adoption, is the next major battleground for core providers.

Nymbus, a modern banking platform for U.S. banks and credit unions, announced the launch of the Nymbus MCP Server, one of the first secure Model Context Protocol (MCP) servers purpose-made for core banking. Built for the Nymbus Banking Platform, the server gives financial institutions a secure, standardized, and controlled way to connect AI-powered experiences to the core through a single interface.

Designed for front-office banking operations, the server currently provides 19 tools that enable AI assistants to support common banking actions such as customer lookup, account management, money movement, and debit card controls. It translates banking capabilities into a format AI agents can securely use to call approved core functions through a large language model, while Nymbus handles the complexity behind the scenes. Financial institutions determine which tools are enabled, which user roles can access them, and where additional review or approval is required.

The result is a practical foundation for financial institutions looking to apply AI to everyday service and operational workflows. For example, a member service agent using an AI assistant can verify a customer, review account details, and initiate an approved debit card freeze workflow through a single conversational interface, without toggling between systems. Similar AI-assisted workflows can help institutions accelerate research-heavy processes such as fraud investigation, case handling, and operational follow-up, improving efficiency while keeping employees in control of the work.

“AI creates real value in banking when it helps institutions get work done, not just generate answers,” said Jeffery Kendall, Chairman and CEO of Nymbus. “With the Nymbus MCP Server, we are giving banks and credit unions a practical way to put AI into everyday workflows while maintaining the control, consistency, and accountability they need to operate with confidence. Our server simplifies common front-office work today while laying the groundwork for more intelligent, agentic banking experiences in the future.”

The Nymbus MCP Server is built on Nymbus’ AI-enabled, open integration architecture and robust analytics layer. Where legacy cores often require custom integrations for every AI tool or use case, the Nymbus MCP Server provides a single, standardized connection layer. That architecture allows institutions to move faster, scale AI-assisted workflows more consistently, and create a stronger foundation for future innovation across internal channels and teams.

Nymbus designed its MCP Server specifically for the regulatory and operational realities of financial services, with token-based authentication, role-based access controls, PII masking in logs, encrypted connections, full audit logging, and granular controls over how institutions choose to deploy and use AI. Together, these capabilities help financial institutions support security, access, and recordkeeping requirements associated with banking compliance programs, while giving them the flexibility to start small, focus on specific use cases, and expand over time within institution-defined guardrails, while keeping data security top of mind. 

Because the server is built on the open MCP standard, it also creates a flexible foundation for the next generation of banking interactions, where AI can move beyond answering questions to assisting employees and coordinating tasks across systems within institution-defined controls.

“Financial institutions are looking at AI as a way to improve service, strengthen operations, and create competitive advantage, but they need a practical path to do that responsibly,” said Matthew Terry, Chief Technology Officer at Nymbus. “The Nymbus MCP Server helps banks and credit unions augment existing processes with AI-assisted workflows that can speed up research, reduce manual effort, and support better decisions, while giving each institution granular control over what is enabled, how it is used, and where governance and auditability are required.”

The Nymbus MCP Server is available as part of the Nymbus Banking Platform. To learn more, go to https://www.nymbus.com/solutions/core/.

FF NEWS TAKE: This is a definitive step in normalizing AI access to sensitive data, effectively moving the needle on operational efficiency in core banking. The industry’s next challenge will be the adoption race. We expect competitors, particularly those serving regional banks and credit unions, to rapidly announce similar, controlled middleware solutions to accelerate AI integration. The true win for Nymbus is standardizing the Model Context Protocol as the secure language between the LLM and the ledger, setting a crucial precedent for auditability and control.

 

StarkWare cuts jobs in reorganization as Starknet revenue plunges 99% from peak

info@uweb3.io
-
0
StarkWare cuts jobs in reorganization as Starknet revenue plunges 99% from peak

StarkWare is restructuring into two business units and cutting staff as it pivots from scaling Ethereum toward building revenue-generating products of its own — a shift forced by a more than 99% collapse in revenue on its flagship Starknet network.

The changes were outlined during a company-wide town hall hosted by CEO Eli Ben-Sasson, where he told employees StarkWare would restructure into two independent units and focus on building revenue-generating products in-house. A transcript of the address to staff was reviewed by CoinDesk.

Starknet chain revenue, which peaked near $6 million in a single month in late 2023, stood at roughly $48,000 through the first half of April 2026, according to DefiLlama data. The decline is partly industry-wide, with Starknet’s competitors equally impacted, as Ethereum’s EIP-4844 upgrade in March 2024 slashed Layer 2 fee revenue across the board.

Total Value Locked (TVL), however, remains above $200 million.

Ben-Sasson told employees the company now needs to “take our technological superiority… and convert it into meaningful revenue, meaningful usage,” signaling a shift away from a pure infrastructure focus toward building products that can drive demand directly.

He added that StarkWare would prioritize building “things that can be done by no other team, in no other way,” focusing resources on products with “immense potential revenue” rather than broad experimentation.

“I started in this field in 2013, almost 13 years ago, and I’ve seen quite a number of winters,” Ben-Sasson said at the town hall. “I think what marks this winter is that there’s a very clear vacuum in leadership across blockchain, and it affects even things like Bitcoin and Ethereum.”

The company will spin up a new revenue-focused Applications unit led by researcher Avihu Levy.

Levy’s promotion comes days after he published a paper outlining Quantum Safe Bitcoin, or QSB, a method for making bitcoin transactions resistant to quantum attacks without requiring changes to the protocol.

The approach replaces traditional signature schemes with hash-based proofs but comes with significant tradeoffs, requiring extensive off-chain computation and costing an estimated $75 to $200 per transaction, versus roughly $0.33 for a standard bitcoin payment.

QSB offers an alternative to BIP-360, a long-pending proposal to add quantum resistance to Bitcoin at the protocol level that was merged to Bitcoin’s improvement proposal repository in February but could take years to activate.

Ben-Sasson did not name Bitcoin or quantum safety as the Applications unit’s target, saying only that StarkWare would focus on products “that cannot be done by any of our competitors” and build with “minimal dependencies on external L1s or external application teams.”

More details, he told staff, would come next week.

A spokesperson for StarkWare declined a request to comment.

BitTensor AI Token Plunges as Top Builder Departs Over Decentralization Doubts

info@uweb3.io
-
0
BitTensor AI Token Plunges as Top Builder Departs Over Decentralization Doubts

In brief

  • TAO has fallen more than 18% in the last 24 hours amid a spat between the network’s founder and a leading ecosystem contributor.
  • The subnet operator, Covenant AI, is leaving BitTensor after alleging centralized control by the project’s founder, Jacob Steeves.
  • Steeves denied the claims, but the market reacted negatively regardless.

Tensions between the founder of BitTensor and a prominent firm building on the decentralized AI network have helped put TAO, the native BitTensor token, into a spiral, falling 18.5% in the last 24 hours amid the public drama. 

The plunge comes as Covenant AI, one of the best-known subnet operators on BitTensor’s network, announced its intentions to leave the ecosystem altogether, alleging malfeasance by BitTensor founder Jacob Steeves, who it claims has acted out against the firm building on his network.

“When a single actor can suspend a subnet’s emissions, override an owner’s authority over their own community spaces, publicly deprecate projects without process, and use token sales as a coercive mechanism to compel compliance, that is not decentralization,” Covenant AI founder Sam Dare posted on X. 

“It is centralized control with decentralized branding,” he added. 

Dare alleged that Steeves suspended Covenant’s subnet emissions, the method by which TAO distributes tokens to miners and validators for performance within subnets. He also alleged that Steeves exerted his control over Covenant’s community spaces, hampering the firm’s ability to communicate with its community. 

But Steeves denied the claims, alleging that it was Dare in fact who was deprecating community channels and deleting posts from within. 

“I do not have the ability to suspend emissions,” Steeves posted on X. 

Covenant recently gained attention for the permissionless training of the Covenant-72B model, an act that was highlighted by billionaire Social Capital founder Chamath Palihapitiya and was detailed to Nvidia founder Jensen Huang on the “All-In Podcast.”

Around the time of the airing in March, TAO surged around 50%, jumping from $247 on March 19 to $370 just a week later. 

The firm operated three subnets—or markets dedicated to producing a specific AI task—on the network. For example, its Templar subnet (SN3) was focused on decentralized pre-training, while Basilica’s (SN39) focus was on decentralized compute—distinct parts of the artificial intelligence stack. 

“We cannot in good conscience continue to build on a network where the foundational claim we make to our investors, that this infrastructure is decentralized and permissionless, is contradicted by the reality of how the network is actually governed,” Dare said in his statement on the incident. 

“It is therefore with deep frustration that we announce Covenant AI’s departure from the BitTensor network,” he added. 

The firm’s trio of subnets now show as “deprecated” among other active subnets according to BitTensor block explorer, Taostats. 

TAO recently changed hands around $272.70, having erased nearly all the gains since Covenant’s model training was noted on the “All-In Podcast.” The token is down about 64% from its all-time high of $757 from May 2024.

Daily Debrief Newsletter

Start every day with the top news stories right now, plus original features, a podcast, videos and more.

Bitcoin hits a wall – the chart just challenged the $88,000 bull case

info@uweb3.io
-
0
Bitcoin hits a wall – the chart just challenged the $88,000 bull case

Just yesterday, CoinDesk reported that despite lingering war risks, analysts foresee bitcoin rallying to $88,000 and higher on the back of several crypto-specific factors, including bullish market flows.

But now, some 16 hours later, the price chart is saying: not so fast.

Bitcoin’s BTC$70,734.80 price has run directly into one of the most pivotal technical levels – a descending trendline that has been in place since October, when Bitcoin peaked above $126,000. And right on cue, the price has turned lower from the trendline resistance.

BTC’s daily chart. (TradingView)

What is a descending trendline and why does it matter

A descending trendline is drawn by connecting a series of progressively lower price peaks over time. Think of a ball dropped from a great height: as it bounces, each rebound is lower than the last. Now imagine linking those lower highs with a straight line – that’s the descending trendline, capturing the fading strength behind each bounce.

In markets, this reflects diminishing buying power, with sellers increasingly asserting control over price action. The longer the trendline holds, and the more often price turns lower from it, the more significant it becomes, signaling a sustained bear phase.

In bitcoin’s case, this particular trendline has been sloping downward since the $126,000 peak in October 2025. That’s roughly six months of lower highs and six months of the market telling you: the trend is down.

This is what traders call a textbook bear market trendline.

The rejection

Since early February, bitcoin has rallied from nearly $60,000 to over $71,000. The sounds bullish on the surface, and in isolation, it is. But zoom out, and you will see immediately that this is a recovery rally within the broader downtrend represented by the descending trendline.

That trendline was tested overnight, and since then, prices have turned lower. This is what aficionados of technical analysis call a trendline rejection, and it means that sellers have overpowered buyers exactly where the bear-market trendline predicted they would.

The market probed resistance, found it, and turned back. Until BTC can close above this trendline on meaningful volume — not just poke through it intraday — this line remains in control, and the broader downtrend remains intact.

Fundamentals tell you what should happen, and analysts on Sunday cited several fundamental datasets, such as Coinbase premium ETF inflows and macro, as catalysts for a rally to $88,000.

However, the price chart tells you what is happening, and right now, the textbook rejection at the six-month bear market trendline is signalling caution for the bulls.

What to Watch From Here

The trendline is the key variable, based on which two scenarios could unfold.

First, the latest rejection at the trendline invites stronger selling pressure, leading ot a deeper decline to $65,000.

The second scenario involves BTC grinding back up, punching through the trendline. That would be a significant positive development, one that would start to align the chart with the bullish fundamental story.

Until the second scenario plays out, the chart and the bull case are telling two different stories.

Aave passes landmark vote ending months-long fight over who controls protocol revenue

info@uweb3.io
-
0
Aave passes landmark vote ending months-long fight over who controls protocol revenue

The governance fight that began when Aave Labs redirected swap fees away from the DAO treasury in December is over, with the community voting in favor of the former.

Aave’s governance has passed the “Aave Will Win” (AWW) proposal, which founder Stani Kulechov called “the most important proposal in Aave’s history” on Sunday, approving a framework that redirects 100% of revenue from all Aave-branded products back to the DAO and consolidates economic rights under a single asset—the AAVE token.

This shift means the DAO is now responsible for funding Aave Labs’ activities. And the proposal passed Sunday approved a $25 million stablecoin grant and 5,000 AAVE token allocation (approximately $6.8 million) to Aave Labs.

Aave DAO, or decentralized autonomous organization, is a governance system that manages the Aave lending protocol, allowing token holders to vote on decisions such as upgrades, fees, and treasury use. It effectively acts as the community-run decision-making body for the protocol.

The vote resolves a dispute CoinDesk reported in December, when delegates flagged that the integration of trading aggregator CoWSwap into Aave’s interface had quietly shifted swap-related fees away from the community treasury to an external recipient.

That controversy exposed a deeper tension over whether Aave Labs or the DAO controlled the protocol’s most valuable asset: its user-facing products and the revenue they generate.

The so-termed ‘Aave Will Win’ (AWW) proposal answers that question decisively in favor of token holders. Protocol revenue, which hit $140 million in 2025 and is tracking to match that in 2026, now gets supplemented by application-layer revenue from Aave Pro, Aave App, Horizon, and Aave Kit. Swaps on Aave.com and Aave Pro are already generating $10 to $20 million in additional revenue on top of existing protocol fees.

“If you own AAVE, you own not just the economic rights of the protocol, but the brand, the users, and the integrations,” Kulechov wrote. Aave Labs commits to working exclusively on Aave-related products under the new structure.

The application layer is where the ambition sits. Aave App will target mainstream users with what Kulechov described as a “fintech-like experience” with $1 million account protection per user and a card launching later that generates fees for the treasury.

The proposal takes a hard line against what Kulechov called “value leakage,” the exact issue that triggered the December dispute. Service providers must build exclusively for Aave, with zero tolerance for relationship gating or products built for themselves at the expense of token holders.

“Payments for posting governance proposals are over,” he wrote.

Every service provider will have measurable goals, and governance process improvements are planned to reduce what Kulechov described as politics and friction.

On the technical side, Aave V4’s reinvestment feature turns idle float capital in lending pools into yield-generating positions, creating an additional revenue stream that did not exist in V3.

New “Spokes” expand collateral options and address the demand side of DeFi liquidity. The team also plans to invest in agentic AI infrastructure for developers building on Aave.

Aave holds roughly $25 billion in total value locked across multiple chains, making it the largest lending protocol in DeFi. The $140 million annual revenue figure puts it alongside Uniswap and Lido as one of the few protocols generating nine-figure income.

Kulechov’s stated target is scaling from $40 billion to $1 trillion, positioning Aave not as a bank but as “a financial network that any fintech, bank, or asset manager can plug into.”

Why Facilities Are Replacing Portable Instruments with Permanent Power Quality Monitoring Systems

info@uweb3.io
-
0
Why Facilities Are Replacing Portable Instruments with Permanent Power Quality Monitoring Systems

Electrical power has always been treated as a given, the assumption being that if the lights are on and the machines are running, everything is fine. That assumption is costing industrial facilities, data centers, and medical environments millions of dollars a year in unexplained equipment failures, voided warranties, process downtime, and regulatory non-compliance. The shift from reactive maintenance to continuous intelligence is no longer optional, and the power quality monitoring system has become the operational standard that serious facility managers are adopting right now. At the centre of that standard is the Inspector V4, a professional-grade power quality analyzer built for permanent 24/7 deployment, not for a week-long site visit, but for the facility that never closes.

The Hidden Cost of Unmonitored Power Quality

Power quality degradation, voltage dips, harmonic distortion, transients, flicker, and frequency variation do not announce themselves. It accumulates silently in the background, shortening the lifespan of motors and drives, corrupting sensitive medical instrumentation, introducing bit errors in data center servers, and tripping protection systems at the worst possible moment. The IEC 61000-4-30 standard exists precisely because regulators recognized that you cannot manage what you cannot measure, and you cannot measure what you only check once a month with a portable instrument pulled from a carry case.

From Event Capture to Real-Time Analysis

The Inspector V4 was designed to solve this problem at the infrastructure level. Operating continuously from inside the facility, it captures every power event, dips, swells, transients, harmonic violations, and energy anomalies, and does not simply record them. It analyses them in real time, applying IEC 61000-4-30 Class A algorithms to produce structured diagnostic data that pushes directly to the Inspector Cloud platform, where engineers access live dashboards, trend reports, and historical analytics from anywhere in the world. When a threshold is crossed, whether a voltage dip exceeds a set limit or a harmonic level breaches compliance, the device fires instant alerts via SMS and Email so the responsible team knows the moment a problem occurs, not the next time someone visits the site. It does not tell you what went wrong last Tuesday. It tells you what is happening right now, why it is happening, and what it means for your equipment.

What Makes a Professional Power Quality Analyzer Different

What separates the Inspector V4 from basic panel meters and entry-level monitors is the depth of its measurement architecture. It carries two fully isolated 3-phase channels, each with five voltage and five current inputs, allowing it to simultaneously monitor two independent electrical systems. Every measurement is captured at 32,000 samples per second through a 24-bit ADC and certified to IEC 61000-4-30 Class A, the highest accuracy class recognized by international regulators, required for contractual power quality reporting and accepted as evidence in legal disputes. A 6-hour lithium-ion battery ensures the device keeps recording even when the power event being measured is the one that caused the outage.

Integration, Cloud Access, and Instant Alerts

Nine industrial communication protocols are built in natively: Modbus, BACnet, IEC 61850, DNP3, MQTT, SNMP, HTTP/HTTPS, IEC 62056, and Email, meaning it integrates directly with virtually any existing SCADA or BMS infrastructure without middleware or third-party gateways. Alongside protocol-based data delivery, the Inspector V4 operates its own cloud infrastructure that aggregates data from multiple installed units, enabling centralised analytics dashboards and multi-site comparison from a single interface. Alarm rules are configured once and executed automatically: when any monitored parameter crosses a defined threshold, the system sends SMS and Email notifications to the designated contacts, ensuring that facility engineers, operations managers, and compliance teams are all notified simultaneously without any manual intervention. Expansion modules further extend the device’s reach: the THPAQ module adds CO₂, TVOC, PM2.5, temperature, humidity, barometric pressure, and 3-axis vibration sensing; the MMU monitors motor and pump health non-invasively; the SIB interfaces up to eight analog sensors across flow, pressure, pH, and temperature.

Who Needs a Permanent Power Quality Analyzer

The applications span every sector where power reliability is operationally critical. Industrial facilities use the Inspector V4 to protect motors, drives, and automation equipment from harmonic-induced stress, with analytics identifying root-cause patterns before they cause production stoppages. In data centers, the device goes well beyond UPS performance monitoring, it maps the entire power path from utility entry to rack level, measures harmonic loading across distribution boards, monitors generator transfer events and transfer time accuracy, tracks power density changes that indicate cooling stress, and catches the subtle voltage irregularities that cause server errors and storage corruption long before any hardware failure is reported. Medical environments rely on it to maintain the strict power quality standards that life-support and diagnostic equipment demand, with Class A accuracy records that satisfy regulatory audit requirements. Telecom operators use it to verify generator and battery backup performance against SLA commitments, while building management teams integrate it directly into BMS platforms to consolidate power, environment, and mechanical monitoring under a single analytics dashboard.

When You Also Need a Portable Power Quality Analyzer

For facilities that also conduct periodic field surveys, commissioning work, or temporary monitoring campaigns at remote assets, Inspector Network offers a complementary solution in the form of its portable power quality analyzer, purpose-built for the engineer who needs to move between sites and gather data on a campaign basis. The two products address fundamentally different operational needs: one stays installed and owns the permanent analytical record of a facility’s electrical health, the other travels where it is needed when it is needed. Together, they cover every power quality monitoring scenario an organization is likely to face, from continuous site-wide intelligence to targeted short-term investigation.

Monitoring as Infrastructure, Not Instrumentation

The economics of permanent power quality monitoring and analytics have always been straightforward. A single prevented equipment failure pays for the monitoring system many times over. What is changing is the intelligence layer now sitting on top of that data. The Inspector V4 does not simply log events and wait for an engineer to interpret them. It analyses continuously, alerts instantly via SMS and Email, feeds structured diagnostic data to a cloud platform accessible from anywhere, and delivers the kind of actionable insight that transforms a passive monitoring installation into an active facility protection system. That is precisely why forward-thinking operators are treating it not as instrumentation, but as infrastructure.

Links in this article:

  1. Homepage, inspector-network.com (anchor: power quality monitoring system) , Paragraph 1
  2. Product page, inspector-network.com/us/product/standard (anchor: power quality analyzer) , Paragraph 1
  3. Portable page, inspector-network.com/us/product/portable (anchor: portable power quality analyzer) , Paragraph 7







South Korea’s Central Bank Pitches Crypto ‘Circuit Breakers’

info@uweb3.io
-
0
South Korea’s Central Bank Pitches Crypto ‘Circuit Breakers’

South Korea’s central bank says crypto exchanges should have their own “circuit breakers” that halt trading to prevent a repeat of the market fallout after Bithumb mistakenly sent more than $40 billion in Bitcoin to its customers in February.

The Bank of Korea said in a payments report on Monday that lawmakers should consider introducing mechanisms similar to the Korea Exchange’s trading curbs to suspend trading if crypto prices suddenly fluctuate.

“Currently, the virtual asset industry lacks internal control mechanisms and faces lower regulatory intensity compared to established financial institutions,” the bank said.

“Consequently, as similar incidents could occur at other virtual asset exchanges, it is necessary to strengthen relevant regulations to prevent them in advance,” the report added.

It comes as South Korean lawmakers are currently looking to pass laws to further regulate crypto, which the Bank of Korea said should include its suggested measures “to enhance the safety and transparency of virtual asset exchange operations.”

In early February, Bithumb erroneously sent customers 620,000 Bitcoin (BTC), worth around $42 billion at the time, instead of 620,000 Korean won, worth $400.

The price of Bitcoin on Bithumb fell as users rushed to sell, causing others to panic-sell and further driving down its price, according to the bank’s report.

A translated graph showing the price of Bitcoin on Bithumb (blue line) compared to Upbit (yellow line) after Bithumb’s erroneous Bitcoin transactions. Source: Bank of Korea

Bithumb halted trading and reversed its Bitcoin sends within minutes, but the exchange said that 1,788 BTC, worth around $125 million, had been sold before it could act, and it covered the shortfall using company reserves.

Related: South Korea tightens crypto withdrawal-delay exemptions after scam losses

The Bank of Korea suggested that crypto exchanges should be required to have systems capable of detecting and preventing “erroneous payments caused by human error.”

It added that exchanges should also have systems to automatically verify a platform’s internal assets compared to those on the blockchain to flag discrepancies.

Magazine: South Korea gets rich from crypto… North Korea gets weapons