Fintech Spotlight

Tag: ai agents

  • Alpic raises $6M pre-seed to launch MCP-native cloud platform for AI agents

    Alpic raises $6M pre-seed to launch MCP-native cloud platform for AI agents

    The Streamroot founders’ new venture aims to provide the infrastructure for AI agents to act directly on digital services.

    Paris-based Alpic has raised $6 million in pre-seed funding to build what it calls the first Model Context Protocol (MCP)-native cloud platform, designed specifically for AI agents. The round was led by Partech, with participation from K5 Global, Irregular Expression, Yellow, Drysdale, Kima Ventures, and Galion.exe, alongside angel investors from companies including Mistral, Datadog, and Dataiku.

    AI models have become adept at generating text and ideas, but enabling them to act—such as booking travel or updating records—has been hampered by workarounds like scraping websites or custom plugins. MCP, a protocol now adopted by major AI players, offers a secure and structured way for agents to connect to external services.

    “Agents need infrastructure built from the ground up, not retrofitted” said co-founder and CEO Pierre-Louis Theron. “The real potential of agents lies in their ability to interact with the digital world around them.”

    Alpic provides a developer platform for deploying and managing MCP servers in minutes, with built-in security, analytics, and tooling. The company says this reduces operational complexity and speeds up production for agent-accessible services.

    The founding team previously built Streamroot, a video delivery startup that helped media companies adapt to streaming. With Alpic, they see parallels between the rise of streaming and the coming shift to agent-first computing.

    “Agent-first protocols like MCP could be as foundational to AI as HTTP was two decades ago to the internet” said Boris Golden, general partner at Partech. “Alpic has the deep infrastructure experience and timing to shape how they will be adopted.”

    After deploying dozens of MCP servers with early customers this summer, Alpic has opened its platform to public beta. The funding will support further product development and scaling as the company positions itself as infrastructure for agent-first computing.

  • Archestra Raises €2.8M for AI Agent Security Platform

    Archestra Raises €2.8M for AI Agent Security Platform

    Archestra has completed a €2.8 million pre-seed funding round in less than two weeks, targeting the enterprise security gap in artificial intelligence agent deployment. The London-based startup’s platform provides oversight mechanisms for companies integrating AI agents with internal data systems.

    The oversubscribed funding round drew participation from Concept Ventures as lead investor, alongside Zero Prime Ventures, Celero Ventures, RTP Global, and Aloniq. Angel investors included Max Hauser from Boston Consulting Group, Nginx co-founder Maxim Konovalov, and executives from ElevenLabs and incident.io.

    Platform Architecture and Enterprise Focus

    Archestra’s solution functions as what the company terms an “MCP orchestrator,” utilizing Anthropic’s Model Context Protocol framework introduced in November. The platform enables Large Language Models to interface with enterprise systems including Slack, email platforms, and human resources software while maintaining security controls.

    “MCP is unlocking a new frontier for AI agents. But, right now, MCP is completely unsuitable for the enterprise” said CEO Matvey Kukuy. The platform addresses concerns about autonomous systems accessing sensitive information inappropriately or operating beyond intended parameters.

    The open-source architecture allows both technical and non-technical users to implement AI agent integrations while preserving compliance requirements and data security protocols.

    Serial Entrepreneur Leadership

    Matvey Kukuy and co-founder Ildar Iskhakov bring prior startup experience to Archestra’s development. The childhood friends previously established incident management platform Amixr, which Grafana acquired in 2021. Following that exit, Kukuy co-founded AIOps platform Keep, later acquired by Elastic.

    Joey Orlando, formerly with Grafana’s engineering team, completes the founding trio as the company’s initial engineering hire. This technical background provides direct experience with enterprise infrastructure challenges that Archestra aims to address.

    Investment Thesis and Market Positioning

    Concept Ventures Principal Ariel Rahamim drew parallels between emerging AI infrastructure and established internet protocols. “Just as APIs became the foundational building blocks for internet infrastructure, Model Context Protocol (MCPs) are emerging as the connective tissue for improving the context layer of AI tools within enterprise” Rahamim explained.

    The investment firm cited the founders’ technical expertise and open-source approach as factors supporting Archestra’s potential market leadership. Integration-platform-as-a-service offerings are projected to generate over $17 billion in revenue by 2028, according to market forecasts referenced by the company.

    Resource Allocation and Growth Strategy

    Archestra plans to direct the new capital toward platform development and team expansion to meet increasing demand for secure AI agent orchestration capabilities. Engineering and product development will receive priority investment as the company scales its open-source offering.

    The rapid fundraising timeline reflects both investor interest in AI infrastructure solutions and the founders’ execution capabilities. Enterprise adoption of AI agents continues accelerating, creating immediate market opportunities for security-focused integration platforms.

    Companies implementing AI agents face the challenge of balancing productivity benefits with data protection requirements. Archestra’s platform provides the governance layer needed for enterprise-scale deployment while maintaining the flexibility that makes AI agents valuable for workflow optimization.

    The company’s focus on the Model Context Protocol positions it within Anthropic’s expanding framework for AI system integration, though the platform supports multiple language model providers through its orchestration capabilities.

  • Refold AI raises $6.5M to automate enterprise integration

    Refold AI raises $6.5M to automate enterprise integration

    Refold AI has secured $6.5 million in combined pre-seed and seed funding to transform how enterprises handle system integration through artificial intelligence. The Bengaluru and San Mateo-based startup aims to eliminate traditional consulting dependencies by deploying autonomous agents that manage complex software integrations.

    AI Agents Replace Traditional Integration Methods

    The platform addresses longstanding inefficiencies in enterprise software integration, a sector historically dominated by consulting firms charging substantial hourly rates. Refold AI’s technology operates through three distinct layers: Workflow Code Agents designed for engineering teams, MCP Chains targeting business users and AI developers, and an Embedded Integrations Platform serving SaaS providers.

    Unlike conventional low-code solutions, the system removes human intervention entirely by embedding senior consultant expertise into automated software agents. These agents independently generate integration workflows, map complex data structures, and perform self-healing when system failures occur.

    Proven Results Across Multiple Industries

    The technology has demonstrated measurable impact across finance, supply chain, and business intelligence applications. Recent implementations include automating reconciliation processes between expense reporting and accounts payable systems, integrating multiple order and inventory platforms to enhance forecasting accuracy, and constructing reverse ETL workflows that supply analytics dashboards with near real-time information.

    Projects that previously required months of consultant time are now completing within days, according to company data. The platform currently serves over 30 enterprises, including Incorta and Naehas, supporting more than 1,500 active users while processing 30 million API calls monthly.

    Financial Performance and Market Position

    Refold AI reports current annual recurring revenue in the low seven figures, with projections targeting $4 million to $5 million ARR by the end of 2025. The company maintains gross margins between 75% and 78%, providing a 24-month operational runway following the recent funding round.

    The startup competes directly with established integration platforms including MuleSoft, Boomi, and Workato, as well as global consulting organizations such as Accenture and Infosys. Refold AI differentiates itself through a code-first approach rather than low-code methodology, utilizing agents to write custom logic contextually while billing based on outcomes rather than billable hours.

    Leadership Team Brings Enterprise Experience

    CEO Jugal Anchalia and co-founder Abhishek Kumar previously sold their startup to Reliance and have extensive experience managing large-scale digital transformations. The founding team’s direct exposure to enterprise resource planning challenges informed their approach to automating integration processes.

    The initial $1.2 million pre-seed round was followed by a $5.3 million seed investment led by Eniac, Tidal, with Better Capital, Ahead VC, Karman Ventures, z21 Ventures and various angel investors.

    Scaling Operations and Future Vision

    The company expects to expand its workforce to 30 employees by year-end as it scales operations to meet growing demand. Refold AI’s stated goal involves becoming the foundational infrastructure that connects high-quality data to AI workflows across enterprise technology stacks.

    “Replace 300,000 consultants with 3,000 lines of code” the company claims, highlighting its ambition to fundamentally alter the integration consulting industry.

    The funding will support continued product development and market expansion as enterprises increasingly seek alternatives to traditional integration approaches. With mounting pressure to reduce consulting costs while improving system reliability, Refold AI’s autonomous approach represents a significant shift toward self-managing enterprise infrastructure.

  • Lendflow unveils AI agent suite for lending operations

    Lendflow unveils AI agent suite for lending operations

    Lendflow has introduced an artificial intelligence automation platform designed to transform lending operations through specialized digital agents. The embedded lending infrastructure provider announced Lendflow Automate at Ai4, positioning the technology as a solution for operational efficiency challenges facing the industry.

    The platform deploys AI-powered agents that handle customer interactions across multiple communication channels while processing documentation and risk assessments. Chief Executive Jon Fry emphasized the scalability benefits during the announcement.

    “Lendflow Automate is about enabling scale without the growing pains” ~ Jon Fry, founder & CEO.

    Multi-Channel AI Agent Architecture

    The automation suite centers on AI operational agents that engage customers through voice calls, text messages, email, and chat interfaces. These agents function continuously, providing round-the-clock support throughout the lending process while learning from each interaction to improve performance.

    Lendflow has implemented the technology internally, reporting savings of more than 500 hours of communication time weekly. The company’s existing customer base has adopted the platform and achieved increased conversion rates alongside reduced operational costs.

    Specialized Communication Agents

    Five primary agents handle different stages of the lending lifecycle. The Application Walkthrough Assistant reconnects with applicants who abandoned their applications, guiding them through completion and document submission. The Schedule Meeting Assistant coordinates appointments between applicants and funding managers through a comprehensive outreach strategy spanning eight days with 47 touchpoints.

    Document Collection Assistant manages required paperwork submission through consistent follow-up communications. The Dead Deal Assistant targets declined or inactive applications with long-term re-engagement strategies. Renewal Outreach Assistant maintains contact with funded merchants over extended periods to identify additional financing opportunities.

    “These AI agents operate like high-performing assistants to your top team members” ~ Jon Fry, founder & CEO.

    Data Processing and Risk Assessment Tools

    Beyond customer communication, the platform includes specialized agents for data classification and risk evaluation. The Industry Map Agent categorizes businesses using NAICS and SIC codes, delivering accuracy rates reaching 100 percent for risk segmentation purposes.

    Doc Analyzer processes financial documents including bank statements and tax returns, reducing manual processing effort by up to 70 percent. The Inbox Automation Agent manages email classification, sender identification, and attachment processing to eliminate manual workflow steps.

    Operational Scale and Implementation

    The complete system operates more than 50 context-aware AI agents designed for specific business functions. Each agent targets measurable outcomes including faster deal closure times and reduced operational expenses. The technology allows human teams to focus on higher-value activities while agents handle routine processes.

    Fry will participate in a panel discussion titled “Leveraging AI to Drive Value” on August 12 at the Ai4 conference, joining representatives from Chime, Mastercard, and Deloitte to discuss artificial intelligence applications in financial services.

    The platform represents Lendflow’s approach to addressing workforce efficiency challenges through artificial intelligence, offering partners what the company describes as an always-available digital workforce for lending operations.

  • Endex secures $14M from OpenAI for excel AI agents

    Endex secures $14M from OpenAI for excel AI agents

    Endex, a startup developing artificial intelligence agents for Microsoft Excel, has secured $14 million in funding led by OpenAI‘s venture fund, creating an intriguing three-way dynamic between the AI company, its biggest corporate backer, and a new player targeting one of Microsoft’s flagship products.

    The San Francisco-based company plans to deploy AI agents directly within Excel spreadsheets to assist financial analysts with data processing, complex calculations, and report generation. Chief Executive Tarun Amasa, a Thiel Fellowship recipient, announced the funding round and product launch simultaneously this week.

    AI Agents Target Financial Workflows

    Endex’s technology leverages OpenAI’s reasoning models to create what the company describes as hour-long autonomous tasks within Excel environments. The AI agents are designed to handle structured analysis work that typically consumes significant time for finance professionals.

    “Finance professionals don’t just need search results; they need structured thinking and deep analysis” Tarun Amasa said in a statement accompanying the funding announcement.

    The startup’s approach focuses specifically on vertical integration within financial workflows, rather than broad-based spreadsheet automation. Amasa’s team spent considerable time working from OpenAI’s San Francisco headquarters throughout the past year, according to social media posts from the CEO.

    Complex Partnership Dynamics

    The investment creates a noteworthy situation where OpenAI backs a company building products for software owned by its primary investor and sometime competitor. Microsoft has invested over $13 billion in OpenAI since 2019, gaining access to the AI company’s intellectual property and resale rights through its Azure cloud platform.

    However, the relationship has grown more complex as OpenAI expands into enterprise services and developer tools that compete with Microsoft’s offerings. The software giant acknowledged this shift by listing OpenAI as a competitor in its 2024 annual report for the first time.

    Market Position and Access

    Endex has begun offering limited early access to users, with Amasa managing distribution through social media engagement on his announcement posts. The company positions itself as serving power users who require advanced analytical capabilities beyond standard spreadsheet functions.

    “We envision a future where every firm has access to teams of digital analysts, seamlessly augmenting time-intensive workflows” Tarun Amasa stated.

    The startup’s launch comes as enterprise software companies increasingly integrate AI agents into existing workplace tools. Endex’s approach differs by focusing exclusively on Excel rather than building standalone platforms or broad office suite integrations.

    Strategic Implications

    OpenAI’s investment in Endex signals continued interest in vertical AI applications, particularly those targeting established software ecosystems. The funding round represents a bet on specialized AI agents rather than general-purpose tools, aligning with broader industry trends toward task-specific automation.

    For Microsoft, the development presents both opportunity and challenge. While Endex could enhance Excel’s capabilities and user engagement, it also demonstrates how external companies can build competitive features using OpenAI’s technology.

    Tarun Amasa, emphasized the collaborative potential: “What excites me most about this collaboration is our shared vision for vertical-specific AI.”