Senior Software Engineer

Plaid gives Sierra's AI agents live access to bank accounts. New Plaid and Sierra integration allows AI agents to access real-time banking data to streamline consumer financing and loan applications. Curated by Financing Your Way from original reporting by PYMNTS. Summary is AI-assisted and editorially reviewed - see its editorial standards. This partnership between Plaid and Sierra is a major step forward for automated consumer financing. It allows AI customer service agents to access real-time banking data during a live chat. For retailers and service providers, this means the 'wall' between a customer asking about financing and actually qualifying for it is disappearing. Instead of an agent telling a customer to go fill out a separate application, the AI can now request account access mid-conversation to verify income or account balances. This technology effectively turns an AI chatbot into a loan officer or financial advisor. When a customer expresses interest in a high-ticket purchase, the AI can pull the necessary data to provide personalized financing offers immediately. This reduces friction in the sales process and prevents customers from dropping off during the transition from your website to a third-party lender portal. By integrating Plaid's secure data fetching with Sierra's conversational AI, businesses can offer instant loan refinancing, credit limit increases, or personalized payment plans without human intervention. For operators, this translates to higher conversion rates on financed purchases and a more seamless digital storefront. Who else is covering this

Strivve Secures Lead Investment from Chartway Ventures in Final Funding Round. Strivve Secures Lead Investment from Chartway Ventures in Final Funding Round marks the latest milestone for the Seattle-based fintech, which announced today that Chartway Ventures will lead what it expects to be its final capital raise. The round also includes existing backers Velera and Reseda Group, completing a CUSO-rooted financing structure that sidesteps traditional venture capital and private equity. Strivve's Top-of-Wallet(R) platform and CardLinks(TM) technology automate card-on-file placement across e-commerce and bill-pay sites, promising up to 96 % success rates for issuers. The new funding will accelerate rollout to additional credit unions and banks, expanding the network of merchants that automatically receive updated card details. What the deal entails. Chartway Ventures, the investment arm of Chartway Credit Union, committed a lead position in a financing round that Strivve describes as its last. While the exact size of the round remains undisclosed, the participation of Velera (formerly PSCU/Co-op Solutions) and Reseda Group signals deep alignment with the credit-union ecosystem. The capital will be deployed to scale the Top-of-Wallet(R) platform, enhance the merchant directory, and broaden integrations with major payment processors. How Strivve's technology works. At its core, Strivve offers a "card-on-file as a service" model. When a cardholder updates a payment card - such as after a renewal or a replacement - the platform pushes the new token to a curated list of over 1,000 merchant and bill-pay sites. The service is embedded in the issuer's digital channels (mobile app, web portal, or email) and leverages a proprietary API that maps card numbers to merchant URLs. The automation eliminates the manual "update-my-card" steps that typically cause cart abandonment. In pilot deployments, issuers have reported a 30 % reduction in failed recurring payments and a 12 % lift in transaction volume attributed to higher card-on-file retention. Why the announcement matters. Stored-card usage is on a steep upward trajectory. Gartner predicts that by 2027, more than 45 % of digital commerce transactions will rely on saved payment credentials, up from 28 % in 2022. Visa's recent data shows guest checkout dropping from 44 % of e-commerce transactions in 2019 to roughly 16 % in FY 2025. In that environment, the ability to keep a card on file without friction becomes a competitive advantage for banks and credit unions seeking to retain spend. Strivve's financing model - backed by credit-union owners rather than Silicon Valley VCs - offers a longer runway and aligns incentives with member-centric outcomes. By avoiding the "fast-money" pressure of typical VC exits, Strivve can focus on sustainable growth, a point highlighted by co-founder David Pool. Competitive context. Strivve competes with a handful of embedded-finance providers that offer token-vault or card-updating services, such as Plaid's Auth API, Marqeta's tokenization suite, and Stripe's Radar for stored cards. Unlike those platforms, Strivve's solution is built specifically for the credit-union ecosystem and integrates directly with CUSO-managed data pipelines. This niche focus yields higher placement success rates - 96 % versus the industry average of 80-85 % reported by Forrester. Moreover, Strivve's merchant directory is continuously refreshed, a capability that rivals like Adobe Commerce Cloud and Salesforce Commerce Cloud rely on third-party data feeds for. The company's ability to push updates in real time aligns with the expectations set by consumer-facing giants such as Amazon and Google, where seamless checkout is the norm. Implications for enterprise Marketing. For enterprise marketers at banks and credit unions, Strivve's platform unlocks a new channel for lifecycle engagement. Automated card updates can be tied to personalized outreach - e.g., email or push notifications that confirm a successful update and suggest related offers. The data generated by successful placements also enriches the member profile, enabling more accurate segmentation in platforms like Microsoft Dynamics 365 or Adobe Experience Cloud. In practice, a credit union could trigger a targeted campaign when a member's card is updated, promoting a new rewards program or a low-interest loan. The higher retention of stored cards directly translates into more reliable recurring revenue streams, a metric increasingly scrutinized by board members in the post-COVID financial landscape. Market landscape. The broader payments infrastructure market is consolidating around three trends: tokenization, open banking, and embedded finance. IDC forecasts a compound annual growth rate (CAGR) of 14 % for token-management services through 2028, driven by regulatory mandates such as Europe's PSD2 and the U.S. Treasury's push for stronger authentication. Within this context, Strivve's CUSO-centric model provides a differentiated path for community-bank and credit-union members who have historically lagged behind larger banks in adopting tokenized payments. By leveraging the collective bargaining power of credit-union networks, Strivve can negotiate bulk merchant onboarding agreements that smaller issuers could not secure individually. The move also signals a broader shift: financial institutions are increasingly looking to internal-focused capital sources - family offices, founder-led funds, and CUSOs - to fund strategic technology initiatives. This trend reduces dependence on public market valuations and aligns product roadmaps with long-term member value rather than short-term exit multiples. Top insights. * Final funding round: Chartway Ventures leads what Strivve calls its last raise, cementing a CUSO-driven capital structure that sidesteps traditional VC pressure. * High success rates: Strivve's Top-of-Wallet(R) platform delivers up to 96 % card-on-file placement, outperforming the industry average of 80-85 %. * Member-centric growth: Automated updates reduce failed recurring payments by 30 % and lift transaction volume by 12 % in pilot programs. * Competitive edge: Built for credit unions, Strivve's solution offers deeper merchant directory integration than generic providers like Plaid or Stripe. * Marketing leverage: Real-time card updates feed richer member data, enabling more precise, automated cross-sell campaigns in CRM suites such as Salesforce and Microsoft Dynamics. * News * August 1, 2026 ThreatMark Named Leader in Behavioral Biometrics & Device Intelligence by QKS Group. ThreatMark Named Leader in Behavioral Biometrics & Device Intelligence by QKS Group - the QKS Group's SPARK Matrix(TM) for 2026 has placed the Pune-based fintech security firm in the top... * News * August 1, 2026 HashKey targets Singapore's APEX in strategic fintech acquisition to consolidate market-infrastructure play. HashKey Holdings Limited (HKEX: 3887) announced that its Singapore-based subsidiary, HKDAG (Singapore) Pte. Ltd., has signed a non-binding framework agreement with the major shareholder of Asia Pacific Exchange Pte. Ltd...

Plaid builds AI model to decode consumer financial behavior. Sequential Foundation Model Gives Banks More Context Beyond Transactions Emilia David - July 29, 2026 Banks and lenders often can't tell the difference between a customer having a bad month and a bad customer having a good month. Financial data company Plaid built a foundation model that it says makes that distinction. Plaid spent about a year building the foundation model, which analyzes how and where money moves and the sequence in which it moves. The sequential foundation model aims to help banks and lenders distinguish customers whose financial lives look identical on paper but manage their money in very different ways. Plaid aims to release the model in either the end of the third or fourth quarter of the year. Suddu Seshadri, head of data and artificial intelligence solutions at Plaid, told ISMG in an interview that the numbers in transactions do not tell the entire picture of how people use their money. "Financial knowledge is very unique because it's not externally available, and you can't necessarily pattern match on text," Seshadri said. "You need to teach the model financial behavior from scratch." Plaid's model looks at each transaction as more than a single isolated data point. Seshadri said it draws on three layers of information: the transaction's meaning, the order and timing of events relative to one another and the account attributes. Together, it allows the model to interpret transactions within the context of the customer's broader financial patterns. Raw transaction data can be too ambiguous and may be understood as just a deposit rather than as salary, severance or reimbursement, which impacts how a person will use that income over time. Since it's too difficult to parse consumer behavior just on how much they spend in a month, Plaid needed to build its own dataset to include not just transactions, but also the ability to correlate where and how often these transactions happen. This is where the sequence of events starts to matter. Plaid uses the example of two people who have the same monthly income, average balance, rent payment, overdraft fees and category-level spending. For consumer A, the moment income arrives, the money goes to pay rent and utilities. Some spending follows with a single overdraft fee for unexpected repairs. The account recovers by the next paycheck and goes back to a normal pattern. Consumer B, on the other hand, goes about things in a different order. Once the paycheck comes in, the money goes to pay loans and credit cards, which drains the account within 24 hours. The consumer makes small transfers right before the next payroll, and then the cycle repeats. Based on the movement of money alone, the two customers may look the same. Still, Seshadri said the sequence of events paints a clearer picture of consumer behavior because it becomes clear that Consumer A does not make it a habit to overdraft their accounts. Plaid can collect this information because of its unique access to the banks it works with and the other companies that connect to it. Seshadri said Plaid trained the model through contrastive learning, which encouraged the model to become better at predicting financial activity, replaced token detection to recognize plausible events and temporal contrastive learning to reinforce behavioral characteristics based on representations from different periods of the user's history. The model trained directly on transaction-level financial behavior, so Plaid said it needed safeguards around how that information is collected, stored and used. Seshadri underscored that, in the process of building its dataset and the sequential foundation model, the company prioritized security and protecting sensitive financial information. The team applied the same security posture that it uses for its entire infrastructure and followed the same security standards as its bank partners - and made sure not to expose any model intervals or thresholds so bad actors don't have an idea of their roadmap. Plaid also got sign-off from its privacy and legal teams before expanding access to the model. Since the model looks at sensitive information that can identify a person's relationship with their money, Seshadri said Plaid focused on ensuring as much of the person's information is secure. The company never accessed information without the individual customer's consent and adhered to existing privacy controls and agreements it made with banks and applications. "When you think about the fraud space, there is a knowledge graph based on behavioral features that are protected, anonymized and de-identified so we see patterns and know if it is unusual," Seshadri said. He added that the data used to train the model, including the movement of money and what the account was used for, was only fully visible to Plaid. Plaid continues to improve the model and add more context to its data. Seshadri said the next goal is to fine-tune the model further and make it easier to use for multiple use cases such as first-party fraud, identity theft, cash advances and payments.