Full-Time

Principal Product Manager

AI Platform

interface.ai

interface.ai

201-500 employees

AI-driven virtual assistants for banking

Compensation Overview

$250k - $320k/yr

San Francisco, CA, USA

In Person

Work is on-site at the San Francisco office.

Category
Product (1)
Required Skills
LLM
Software Testing
Product Management
Machine Learning
RAG
SOC 2
Cryptography
REST APIs
DevOps

Get referred to interface.ai

See people who can refer or advise you

Requirements
  • At least 8 years of product management experience, including at least 3 years at Staff or Principal level on platform, artificial intelligence/machine learning, or infrastructure products.
  • Experience owning a multi-component platform serving multiple product teams.
  • Deep technical fluency, including the ability to read code, review architecture, and debug production issues alongside engineers.
  • Experience with artificial intelligence or machine learning platforms, including large language model orchestration, model evaluation, fine-tuning, retrieval-augmented generation, agentic frameworks, or similar systems.
  • Experience governing platform contracts and managing how internal teams consume platform capabilities.
  • Experience building in enterprise software-as-a-service environments with multi-tenant, compliance, and governance requirements.
  • Ability to build with artificial intelligence daily by creating agents, automating analyses, prototyping with code, and building custom workflows.
Responsibilities
  • Own the artificial intelligence platform that supports every product at the company.
  • Assess the existing multi-component platform and define what to consolidate, invest in, and sunset.
  • Set technical direction for the platform pillar and partner with engineering leadership to scale the system to millions of end users.
  • Define agent runtime and orchestration capabilities, including multi-agent orchestration, memory and context management, tool routing, goal planning, and policy-aligned behavior.
  • Define platform application programming interface contracts, versioning, deprecation, and migration strategy for consuming teams.
  • Own model lifecycle decisions, including model selection, evaluation, benchmarking, fine-tuning enablement, and prompt optimization.
  • Manage inference performance and reliability through latency budgets, caching, batching, quantization, multi-region deployments, fallbacks, canary releases, feature flags, and rollback patterns.
  • Own safety and compliance capabilities, including personally identifiable information detection and masking, prompt and response safety policies, tenant isolation, SOC 2 and General Data Protection Regulation readiness, audit logs, and encryption.
  • Define evaluation and quality systems, including offline golden sets, regression testing, hallucination measurement, task-success metrics, and production-to-improvement feedback loops.
  • Define voice and multimodal integration patterns for automatic speech recognition and text-to-speech, including word error rate, interruption handling, turn-taking, and real-time latency budgets.
  • Mentor product managers, raise the quality bar, provide input on hiring, prioritize competing demands across product pillars, define the platform investment thesis, and protect the platform from becoming a feature factory.
  • Build tools, agents, scripts, automated analyses, prototypes, and workflows to accelerate product work.
Desired Qualifications
  • Experience in financial services or another regulated industry.
  • Experience with voice, automatic speech recognition, or text-to-speech systems and real-time latency-sensitive products.
  • Experience with large language model evaluation frameworks, multi-agent orchestration, reinforcement learning from human feedback, or reinforcement learning from artificial intelligence feedback.
  • Experience with knowledge graphs or retrieval-augmented-generation retrieval systems.
  • Experience building a platform team from scratch.

Interface.ai provides an AI-powered platform that delivers intelligent virtual assistants for banks and credit unions. Its agents operate across voice, chat, and mobile as a 24/7 personal bank teller, with Agentic Voice AI, Agentic Chat AI, and Agentic Employee AI, and are pre-trained on 60+ banking use cases while integrating with core banking, CRM, and LOS systems. It differentiates itself by targeting community banks and credit unions with ready-to-use, industry-specific AI and end-to-end automation, backed by a performance-based pricing model. Its goal is to improve operational efficiency, boost customer self-service, strengthen regulatory compliance, and help financial institutions scale AI-powered service across channels.

Company Size

201-500

Company Stage

Early VC

Total Funding

$80M

Headquarters

San Francisco, California

Founded

2018

Get referred to interface.ai

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • Smart Collections launched February 19, 2026 targets delinquency when credit union defaults worsen.
  • Red Rocks and Navigator show live production traction across voice automation and member service.
  • 20 million dollars from Avataar in 2024 plus 100-plus customers funds faster product expansion.

What critics are saying

  • A Hallucination or compliance failure in lending, collections, or authentication destroys trust instantly.
  • Banks can switch to Five9, RingCentral, or internal copilots once CCaaS bundling commoditizes.
  • If Nexus underperforms pilots, interface.ai remains a niche vendor serving under 100 institutions.

What makes interface.ai unique

  • BankGPT, launched November 18, 2025, is banking-specific and action-oriented, not generic chat.
  • Telarus partnership on February 18, 2026 packages elite CCaaS with banking-native AI.
  • Nexus on April 14, 2026 attacks hold queues with micro-request routing and no transfers.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Health Insurance

Paid Vacation

401(k) Retirement Plan

Parental Leave

Professional Development Budget

Growth & Insights and Company News

Headcount

6 month growth

1%

1 year growth

3%

2 year growth

27%
PR Newswire
May 4th, 2026
AI Interfaces, Inc. announces launch of KongXLM(TM) and OMNiEYE(TM) financial Prediction Engine.

AI Interfaces, Inc. announces launch of KongXLM(TM) and OMNiEYE(TM) financial Prediction Engine. May 04, 2026, 08:37 ET Multi-Model AI Orchestration Platform and Swarm Prediction Engine Now Available in Public Beta SAN FRANCISCO, May 4, 2026 /PRNewswire/ - AI Interfaces, Inc. today announced the public launch of KongXLM(TM) and OMNiEYE(TM), a unified platform designed to address fragmentation and opacity in modern artificial intelligence systems. The platform is now available in beta at app.kongxlm.com. "Artificial intelligence today is powerful but fragmented. Every model thinks differently, and that fragmentation creates both confusion and opportunity," said Rob Shambro, Founder of AI Interfaces, Inc. "KongXLM(TM) was built to orchestrate that intelligence, not replace it. OMNiEYE(TM) extends that foundation into real-world decision-making by turning multi-model reasoning into structured, high-confidence predictions." KongXLM(TM): Multi-Model AI Orchestration KongXLM(TM) enables users to submit a single prompt across 21 leading AI models simultaneously. Outputs are presented in parallel, allowing direct comparison of reasoning, tone, and conclusions. The platform emphasizes divergence as a signal and consensus as a confidence indicator. OMNiEYE(TM): Parallel AI Prediction Engine OMNiEYE(TM) executes large-scale parallel analysis using 30 specialized agents across multiple models. Each prediction generates hundreds of independent analytical outputs, synthesized into a structured result including directional outlook, confidence score, and agreement metrics. 21+ teams create a 630 Agent swarm, leading to a debate, prediction and level of confidence, a 9 step process. Portfolio Labs: Scenario Modeling Portfolio Lab enables simulation of macro and event-driven scenarios, dynamically repricing positions and providing structured recommendations including hold, adjust, or exit strategies. Availability and Demonstrations All features are available through a free trial until May 20, 2026. KongXLM(TM): app.kongxlm.com/demo-orchestration OMNiEYE(TM): app.kongxlm.com/demo-omnieye Portfolio Labs: app.kongxlm.com/portfolio-lab-demo About AI Interfaces, Inc. Founded in 2025 by Rob Shambro, AI Interfaces, Inc. is building the control plane of artificial intelligence. The company develops multi-model orchestration systems that sit above leading AI models, enabling unified access, comparative reasoning, and large-scale decision intelligence through a single interface. Its core platform, KongXLM(TM), orchestrates and synthesizes outputs across multiple AI systems, turning model disagreement into actionable intelligence, while OMNiEYE(TM) extends that architecture into predictive analytics and financial decision-making. Combined with Portfolio Labs, the company is defining a new category of AI infrastructure focused on orchestration, prediction, and real-world execution. AI Interfaces, Inc. products are built for builders, traders, and enterprises who require more than a single model's perspective. The company is headquartered in St. Petersburg, Florida, and operates globally at app.kongxlm.com. Media Contact AI Interfaces, Inc. 7901 4th St N #14642 St. Petersburg, FL 33702, USA Email: [email protected] Website: www.kongxlm.com App: https://app.kongxlm.com SOURCE AI Interfaces, Inc.

GlobeNewswire
Apr 14th, 2026
interface.ai launches Nexus - The world's first fully agentic CCaaS platform that ends the era of hold queues, transfers, and binary call routing for credit unions and community banks.

interface.ai launches Nexus - The world's first fully agentic CCaaS platform that ends the era of hold queues, transfers, and binary call routing for credit unions and community banks. April 14, 2026 13:04 ET | Source: interface.ai New platform shatters the decades-old "AI handles it or a human handles it" model - introducing a third paradigm where AI owns the member conversation end-to-end while human judgment arrives in seconds, not minutes - multiplying agent capacity up to 10x and eliminating hold times entirely. Announced at Financial Brand Forum 2026, Booth B403. PALO ALTO, Calif., April 14, 2026 (GLOBE NEWSWIRE) - interface.ai, the definitive leader in agentic AI for community financial institutions, today unveiled Nexus at Financial Brand Forum 2026 - the industry's first fully agentic Contact Center as a Service (CCaaS) platform purpose-built for credit unions and community banks. Nexus eliminates the foundational constraint that has defined contact centers for decades: the binary choice between AI handling a call or transferring it to a human. With Nexus, AI owns the conversation while human expertise is available in seconds - no transfers, no hold queues, no repeated information - fundamentally rewriting the economics of every contact center that deploys it. A Third Paradigm: AI Owns the Conversation. Humans Provide Judgment. The contact center industry has operated on a binary model since its inception. When AI or an IVR cannot handle a request, the entire call is transferred to a human agent. The member waits on hold. The agent rebuilds context from scratch. A request that required 10 seconds of human judgment consumes 4 minutes of agent time. Every CCaaS vendor on the market today - from legacy incumbents to AI-native startups - operates within this same binary framework. They optimize within the constraint. They do not eliminate it. Nexus eliminates it. Nexus introduces a fundamentally new architecture for how contact centers operate. The AI does not hand off the call. It does not transfer the member. It does not put anyone on hold. Instead, when a moment in the conversation requires human judgment - an approval, a policy exception, a verification - Nexus routes only that specific micro-request to the next available agent with the right skill. The agent resolves it in seconds. Nexus delivers the answer back to the member at a natural moment in the conversation, never interrupting mid-sentence. The member hears one continuous, seamless experience. Members can make multiple requests in a single call. Each is routed independently to the right skilled agent. Two agents, three agents working in parallel - the member hears one voice, one conversation, zero transfers. Two Agentic Modes Built for the Reality of Community Finance Nexus operates in two modes that address different dimensions of the contact center challenge: Agentic Mode: AI leads, agents assist in real time. Nexus AI manages the full member conversation across voice and chat. When human judgment is needed, it escalates only the specific request into skills-based queues. The right agent sees it, resolves it with a single action, and Nexus seamlessly delivers the response. No hold. No transfer. No context rebuilding. Multiple requests, multiple agents, one unbroken conversation. Co-Pilot Mode: Agents lead, Nexus accelerates everything around them. When a human agent is already in conversation, Nexus automates the work that consumes the most time. For loan applications and account openings, Nexus sends secure document collection links, auto-extracts data, and pre-fills forms - compressing 25 minutes of manual data entry into 2. If documents are not available, Nexus fills applications from live conversation transcription as the member speaks. In both modes, Nexus shows UI to agents only when their attention is required. No monitoring dashboards. No idle screens. Agents see work when it exists and are free when it does not. Why This Changes the Economics of Community Banking The implications for credit unions and community banks are transformative: Agent capacity multiplied, not merely optimized. When agents resolve micro-requests in seconds rather than managing entire calls for minutes, each agent effectively handles 5-10x more member interactions. Institutions scale capacity without scaling headcount. Hold times eliminated, not reduced. Members never wait in a silent queue. When all agents are busy, Nexus continues the conversation - gathering context, verifying identity, resolving what it can, and delivering personalized product recommendations through integration with Intelligent Banking. Wait time becomes productive time. Service time becomes revenue time. New agents are productive immediately. Nexus embeds institutional knowledge - processes, policies, procedures, disclosures - directly into the workflow. Quick actions tell agents exactly what to do. Forms are pre-filled. Compliance language is presented on screen. The weeks of training that traditionally stand between a new hire and productivity are compressed to the first interaction. Revenue-aligned routing. Every request enters skill-based queues with configurable priorities. Institutions can prioritize loan applications and new account openings over routine inquiries - directly connecting contact center operations to revenue and membership growth objectives. Integrates with your current CCaaS - or replaces it entirely. Nexus layers directly onto existing CCaaS platforms through native integration, giving institutions an immediate agentic upgrade without disrupting current infrastructure. Or, institutions can route calls directly to Nexus and replace their legacy CCaaS altogether - running a fully agentic contact center end-to-end on a single platform purpose-built for community finance. "Contact centers have been trapped in a binary model for decades: either the AI handles it, or a human handles it. The entire industry has optimized within that constraint without questioning it," said Srinivas Njay, Founder and CEO of interface.ai. "Nexus introduces a third option that should have existed years ago. The AI owns the conversation while human judgment arrives in seconds - no transfers, no holds, no members repeating themselves. This is not an upgrade to call routing. It is the end of call routing as the industry knows it. And for credit unions that deploy it, the economics change overnight." Powered by BankGPT - The Proven Foundation for Agentic Banking Nexus is built on interface.ai's BankGPT platform, the industry's first agentic AI platform for community finance, trusted by 100+ financial institutions and operating at over 1.5 million conversations per day across banking use cases. BankGPT provides the banking-grade conversational intelligence, centralized governance and compliance controls, and continuous learning loops that make Nexus possible - and that no horizontal CCaaS or generic AI provider can replicate. Nexus is available immediately for pilots and phased rollouts across voice and chat channels for credit unions and community banks. Institutions interested in early access can contact interface.ai for a live demonstration. About interface.ai interface.ai builds specialized AI agents for credit unions and community banks to help them deliver faster service, stronger relationships, and better outcomes - at scale. The company's BankGPT platform powers agentic AI across voice and digital channels and is trusted by 100+ community financial institutions. With 99.99% uptime, more than 1.5 million conversations per day, and 500+ million conversations to date, interface.ai provides a proven foundation for community-focused banking AI. To learn more, visit interface.ai.

TechAnnouncer
Mar 21st, 2026
Interface.ai raises $20M to expand AI banking platform handling 60% of customer inquiries

Interface.ai, an AI platform provider for banks and credit unions, has raised $20 million in its first external funding round, led by Avataar Venture Partners. The company, which has been self-funded since its 2019 founding, serves over 100 financial institutions. The funding will enhance Interface.ai's proprietary generative AI technologies and expand its team to accelerate product development. The company's AI agents currently manage over 60% of initial customer enquiries using advanced natural language processing to deliver personalised banking experiences. Interface.ai plans to strengthen its regulatory compliance capabilities and speed up go-to-market strategies through increased sales, marketing and partnership development. The company aims to become a billion-dollar revenue business by scaling its AI solutions across voice, chat and branch operations.

Associated Press
Feb 19th, 2026
Interface.ai launches Smart Collections AI agent for credit unions and community banks to tackle rising delinquency

Interface.ai has launched Smart Collections, a multi-channel AI agent designed to help credit unions and community banks manage early-stage delinquency. The announcement comes as credit unions reported total delinquency of 95 basis points in Q3 2025, with 30-59 days past due delinquencies rising to 1.13% in September 2025. Smart Collections combines outbound voice, two-way SMS and email AI agents to handle collections across various products including overdrafts, credit cards, auto loans and mortgages. The system features policy controls, journey orchestration and automated quality assurance to ensure compliance whilst maintaining member relationships. Built on interface.ai's BankGPT platform, the solution integrates with core banking systems from Jack Henry, FIS, Fiserv and others. Interface.ai currently serves nearly 100 community financial institutions, processing over 1.5 million conversations daily.

Yahoo Finance
Feb 19th, 2026
interface.ai Launches Smart Collections - An Agentic, Multi-Channel Collections AI Agent Built for Credit Unions and Community Banks to Increase Payments

interface.ai launches Smart Collections - an agentic, multi-channel collections AI Agent built for credit unions and community banks to increase payments. interface.ai New solution automates early-delinquency outreach across voice, SMS, and email and increases payments - combining product-aware playbooks, policy-gated compliance, and self-service controls to improve cure rates while protecting member relationships PALO ALTO, Calif., Feb. 19, 2026 (GLOBE NEWSWIRE) - interface.ai, the leading provider of specialized AI agents for credit unions and community banks, today announced the launch of Smart Collections - a new multi-channel, agentic collections AI designed to help community financial institutions cure delinquency earlier, before accounts roll to later stages and recovery costs rise. Early-stage delinquency is becoming a sharper focus across the industry. Credit unions reported total delinquency of 95 basis points in Q3 2025, and consumer credit data showed 30-59 days past due delinquencies rising to 1.13% in September 2025, nearing pre-pandemic levels - underscoring why earlier engagement and higher-quality outreach matter now more than ever. Built for the reality of community finance: limited staffing, high service expectations, tight compliance Collections leaders at credit unions and community banks are under constant pressure to do more with less - without sacrificing member trust. Smart Collections addresses this with coordinated automated outreach and compliance-by-design: * Outbound Voice AI Agent that can place calls, verify identity (policy-gated), capture Promise-to-Pay, send payment links, and escalate to a human when needed * Two-way SMS AI Agent that supports STOP/opt-out handling, payment link flows, intent detection (pay, dispute, hardship), and compliant cadence controls * Two-way Email AI Agent with deliverability safeguards, templates, reply handling, and preference management * Product-aware playbooks for the products most common in CU/community bank portfolios: * negative balance / overdraft, overdraft line of credit * credit card, auto/RV/boat, personal installment, personal LOC * share/CD-secured credit builder loans * HELOC/home equity * optional mortgage early-intervention workflow with timing gates and notices (where applicable) Differentiated from "point solutions" - comprehensive agentic execution with auditable controls Smart Collections is designed for institutions that want more than a standalone collections workflow system, a digital self-cure portal, or a generic omni-channel messaging tool. Many solutions in the market focus on one layer - collections case management and analytics, a borrower self-service portal, or communications routing - leaving teams to stitch together policies, orchestration, AI behavior, and QA.