API World + CloudX + DataWeek 2025: Full Schedule

Mufaddal Khambaty, Pax8, Senior PM, Integrations & AI

Many platforms start their API journey by issuing simple API keys, but quickly run into security and visibility challenges, especially when third-party developers enter the mix. This session details Pax8's evolution from insecure, direct API key usage to building a sophisticated, secure, and scalable 'Integrations Hub.' We'll share our real-world experience tackling the initial problem: how to allow third-party vendors to access customer data via APIs without compromising security or losing track of who was making calls.

Speakers

Mufaddal Khambaty

Senior PM, Integrations & AI, Pax8

Entrepreneurial and product background. Ex-product leader at Zillow, Custom Ink, Accurate Background. Started as a strategy consultant.

Friday September 12, 2025 9:30am - 9:55am PDT
VIRTUAL API World -- Workshop Stage A (PRO)

API World

Talk Type PRO Session
Tracks API World, Partnership and Integration Strategy (API World), API Program and Developer Relations (API World), Virtual
In-Person/Virtual Virtual

9:30am PDT

[Virtual] PRO Session (CloudX): Turning Tech Debt into Impact: A SRE's Guide to Prioritization and Action

Friday September 12, 2025 9:30am - 9:55am PDT

VIRTUAL CloudX -- Main Stage

Saurabh Phaltane, Google, Senior Site Reliability Engineer

My talk will delve into my experiences managing technical debt in a large-scale search environment, specifically how I've proactively used my SRE expertise to turn what's often seen as a burden into a source of impactful projects. I'll walk through the key principles that guide my approach, including how I categorize and prioritize tech debt, and how I collaborate with other teams to address it effectively. A core part of my talk will focus on the methods I employ to pinpoint high-impact opportunities within the landscape of existing tech debt – looking at pain points, identifying patterns, and always considering the business impact.
Finally, I'll emphasize the importance of "cleaning up the mess" as a vital aspect of SRE hygiene, discussing how automation, monitoring, and thorough documentation contribute to long-term system health and prevent the resurgence of technical debt. I'll illustrate these points with concrete examples from my work, demonstrating how addressing tech debt can lead to significant improvements in system performance, reliability, and overall engineering efficiency.

Speakers

Saurabh Phaltane

Senior Site Reliability Engineer, Google

Saurabh Phaltane, a seasoned Senior SRE, delivers resilient and scalable infrastructure solutions. His deep understanding of large-scale optimization, driven by attention to detail and automation, directly improves system performance and reliability in complex environments. He is... Read More →

Friday September 12, 2025 9:30am - 9:55am PDT
VIRTUAL CloudX -- Main Stage

CloudX

Talk Type PRO Session
Tracks CloudX, Managing AI/ML in the Cloud (CloudX), Cloud Native Development (CloudX), Virtual
In-Person/Virtual Virtual

9:30am PDT

[Virtual] PRO Session (DataWeek):The Rise of Agentic Commerce: Where AI Intelligence Meets Infinite Retail Possibility

Friday September 12, 2025 9:30am - 9:55am PDT

VIRTUAL DataWeek -- Main Stage

Aswini Atibudhi, Walmart, Driving Innovation with Generative AI

In this visionary session, we’ll explore how Agentic AI is not just enhancing retail — it’s fundamentally reinventing it. As commerce shifts from static transactions to dynamic, intelligent interactions, Agentic AI emerges as the architect of a new era: Agentic Commerce. Learn how AI-powered agents, composable architectures, and autonomous decision-making systems are creating infinitely adaptable, customer-first retail ecosystems. We’ll dive into real-world examples, transformative architectures, and the strategic shifts needed to thrive in an AI-first retail future.
Join us to uncover how businesses can unlock limitless innovation, personalized experiences, and operational agility by embracing the rise of Agentic Commerce.

Speakers

Aswini Atibudhi

Distinguished Architect for Customer Space, Walmart

Aswini is the Distinguished Architect for Customer Space at Walmart. He has more than 20 years of IT experience in design and development of scalable microservice and microfrontend based Web/Cloud/AI & ML Applications. His Portfolio is jam-packed with Multiple domains such as Finance... Read More →

Friday September 12, 2025 9:30am - 9:55am PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type PRO Session
Tracks DataWeek, AI Models and Management (DataWeek), Data Tools / Technology and Management (DataWeek), Virtual
In-Person/Virtual Virtual

10:00am PDT

[Virtual] PRO Session (CloudX): Cloud-Native AI/ML: Deployment, Scaling, and Optimization

Friday September 12, 2025 10:00am - 10:25am PDT

VIRTUAL CloudX -- Main Stage

Prashanth Lakshmi Narayana Chaitanya Josyula, Salesforce, Principal Member of Technical Staff
Anant Kumar, Salesforce, Technical Leader

The rapid evolution of artificial intelligence and machine learning is transforming how cloud-native applications are built and managed. This session will explore key strategies for deploying AI/ML workloads in the cloud, ensuring scalability, and optimizing performance. We’ll discuss best practices for leveraging Kubernetes, Kubeflow, and serverless architectures, along with automation techniques for model training, deployment, and monitoring. Attendees will gain insights into cost-effective scaling, security considerations, and the future trends shaping AI/ML in cloud environments. Whether you’re an ML engineer, cloud architect, or DevOps professional, this session will equip you with practical knowledge to enhance your AI/ML workflows.

Speakers

Anant Kumar

Technical Leader, Saleforce

Lead Member of Technical Staff at Salesforce, I currently lead the Big Data Compute Services team within the Einstein AI platform, where I architect and manage data lake solutions that provide critical APIs for AI model training. Our platform powers Salesforce's predictive and generative... Read More →

Prashanth Lakshmi Narayana Chaitanya Josyula

Principal Member of Technical Staff, Salesforce

Hello! I’m Prashanth, an avid connoisseur of technical literature and a dedicated practitioner in the field of programming. With a deep-seated passion for technology, I excel in crafting sophisticated code and exploring cutting-edge innovations.Currently serving as a Principal Member... Read More →

Friday September 12, 2025 10:00am - 10:25am PDT
VIRTUAL CloudX -- Main Stage

CloudX

Talk Type PRO Session
Tracks CloudX, Managing AI/ML in the Cloud (CloudX), Cloud Native Development (CloudX), Virtual
In-Person/Virtual Virtual

10:00am PDT

[Virtual] PRO Session (DataWeek): Data Sovereignty in the Age of AI

Friday September 12, 2025 10:00am - 10:25am PDT

VIRTUAL DataWeek -- Main Stage

Michel Tricot, Airbyte, Co-founder and CEO

This session will explore the intersection of data sovereignty and artificial intelligence, addressing how organizations can maintain control of their valuable data assets while still leveraging the power of AI. Drawing from extensive experience building open-source data infrastructure solutions, Michel will illuminate the challenges companies face when integrating AI into their data ecosystems without compromising ownership, security, or compliance requirements.

The session targets data leaders, CDOs, and enterprise architects who are navigating the complex landscape of AI adoption while maintaining strict data governance standards. Michel will share practical frameworks for implementing a self-managed data integration strategy that enables AI innovation while preserving first-party data sovereignty—a crucial consideration as regulatory requirements around data protection continue to evolve globally. Attendees will gain actionable insights on building resilient data architectures that support AI initiatives without surrendering control of sensitive information.

This session aligns perfectly with Data Week's focus on "Data Engineering & Governance" and "AI & ML" tracks, offering attendees a unique perspective on balancing innovation with control. Conference participants will benefit from Michel's vision of how open-source data integration infrastructure can serve as the foundation for responsible AI development, empowering organizations to build competitive advantages while maintaining complete sovereignty over their data. The presentation will include real-world examples of companies that have successfully implemented these principles.

Speakers

Michel Tricot

Co-founder and CEO, Airbyte

Michel Tricot is co-founder and CEO of Airbyte, the open data movement platform. The company was started in 2020 with a vision of commoditizing data integration pipelines across all industries and organizations and today has more than 170,000 deployments. Michel has been working in... Read More →

Friday September 12, 2025 10:00am - 10:25am PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type PRO Session
Tracks DataWeek, AI Models and Management (DataWeek), Data Engineering / Architecture and Streaming (DataWeek), Virtual
In-Person/Virtual Virtual

10:30am PDT

[Virtual] OPEN Session (CloudX): A Design Delivery System For AI and Overcoming Tech Debt

Friday September 12, 2025 10:30am - 10:55am PDT

VIRTUAL CloudX -- Main Stage

Megha Sandesh, Solvedex, CEO

Most AI projects don’t fail because of bad ideas—they fail because of bad systems. Teams rush to deploy models without a clear design framework, leading to disjointed interfaces, brittle integrations, and mounting technical debt that becomes impossible to untangle. The result? AI that never reaches its full potential, burdened by rework, inconsistency, and scalability nightmares.

My session breaks down a pragmatic, battle-tested approach to AI design and delivery—one that isn’t just about getting a model into production but ensuring it thrives there. We’ll dig into how design, engineering, and operations must work in sync from day one to prevent the common pitfalls that derail AI initiatives.

Key takeaways include:

1. Why AI projects accumulate debt faster than traditional software—and how to stop it at the source.
2. The role of system design in AI, and why model accuracy alone won’t save a broken pipeline.
3. A real-world approach to designing AI that scales without accumulating chaos.

Speakers

Megha Sandesh

CEO, Solvedex & SolveX Design

Visionary leader with extensive experience in scaling companies and leading cross-functional teams to deliver exceptional results.Built a successful bootstrapped company from ground zero to $5M ARR in under 2 years through innovative lean strategies, cost optimization, and a relentless... Read More →

Friday September 12, 2025 10:30am - 10:55am PDT
VIRTUAL CloudX -- Main Stage

CloudX

Talk Type OPEN Session
Tracks CloudX, OPEN Session, IT Operations and Enterprise Management (CloudX), Leading a Cloud-First Company (CloudX), Virtual
In-Person/Virtual Virtual

10:30am PDT

[Virtual] OPEN Session (DataWeek): Ask Us Anything: Building Real-World AI Agents for Business

Friday September 12, 2025 10:30am - 10:55am PDT

VIRTUAL DataWeek -- Main Stage

Lee Dickson, raia AI, Director of Sales and Operations
Rich Swier, raia AI, Founder

You’ve seen the hype. Now let’s get practical. In this live AMA-style session, Lee Dickson and Rich Swier — co-hosts of The AI Guys podcast and builders of raia AI — open the floor to the questions business leaders, operators, and tech pros are really asking about AI.

From deployment timelines to hallucination prevention, ethics to employee augmentation — nothing’s off limits. Whether you’re planning your first AI rollout or scaling from one use case to a dozen, this is your chance to get direct insights from AI practitioners who’ve helped companies from $5M startups to $500M enterprises make AI work.

Speakers

Lee Dickson

Director of Sales and Operations, raia AI

Lee Dickson brings 7+ years of experience in productizing AI solutions and predictive analytics, with a strong track record of streamlining operations and improving customer engagement for SMBs and enterprise clients. With an extensive background in technology and SaaS, he's an advocate... Read More →

Rich Swier

Founder, raia

Rich is a serial entrepreneur based in Sarasota, Florida. For the past 30 years, Rich has built and exited numerous successful tech businesses and continues to launch and incubate new ventures.

Friday September 12, 2025 10:30am - 10:55am PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type OPEN Session
Tracks DataWeek, OPEN Session, Data Tools / Technology and Management (DataWeek), AI Models and Management (DataWeek), Virtual
In-Person/Virtual Virtual

11:00am PDT

[Virtual] OPEN Session (CloudX): Conquering Data Consistency Challenges in Cloud-Native Applications

Friday September 12, 2025 11:00am - 11:25am PDT

VIRTUAL CloudX -- Main Stage

Brijesh Deo, Oracle, Software Development
Todd Little, Oracle, Chief Architect

Data consistency is critical in the era of cloud-native applications built on microservices architecture due to distributed services and data stores. Inconsistent data can lead to incorrect decision-making, financial losses, downtime, and loss of reputation. Transactions are the proven method to ensure data consistency and can be effectively implemented with cloud-native applications. In this session, you will learn how to overcome the data consistency challenges by employing various transaction patterns. We will demonstrate how a Spring Boot application can, with minimal code changes, use these transaction patterns to ensure data consistency while maintaining scalability & reliability. We will also discuss how similar applications can be developed with other frameworks, such as Express.js, Micronaut, and Helidon, etc.

Speakers

Brijesh Deo

Director, Software Development, Oracle

Brijesh Deo is the Director of MicroTx product development in the database organization at Oracle. He is passionate about distributed systems and cloud technologies and specializes in Kubernetes, Microservices, and Java. Brijesh has a rich industry experience of more than two decades... Read More →

Todd Little

Chief Architect, Oracle

Todd Little is chief architect for transaction processing products at Oracle, including the Oracle Tuxedo product family, Oracle Blockchain Platform, and Oracle Transaction Manager for Microservices (MicroTx). Todd is responsible for defining the technical strategy and direction for... Read More →

Friday September 12, 2025 11:00am - 11:25am PDT
VIRTUAL CloudX -- Main Stage

CloudX

Talk Type OPEN Session
Tracks CloudX, OPEN Session, Cloud Native Development (CloudX), Virtual
In-Person/Virtual Virtual

11:00am PDT

[Virtual] OPEN Session (DataWeek): Data Integrity in the Age of AI: SBOMs, Lineage, and Trust in the Pipeline

Friday September 12, 2025 11:00am - 11:25am PDT

VIRTUAL DataWeek -- Main Stage

Saloni Garg, Wayfair, Senior Software Engineer

With AI models consuming more data than ever, ensuring the integrity and traceability of that data is critical. This talk focuses on how to build trust into your data pipelines -- using concepts like SBOMs (Software Bill of Materials) for datasets, audit trails, and metadata tagging to make data consumption safer and more transparent. I’ll also touch on how this ties into emerging compliance frameworks and how we’ve approached this in practice.

Speakers

Saloni Garg

Senior Software Engineer, Wayfair

International Red Hat Women in Open Source Awardee | Mozilla Open Leader 2019 | a strong open source diversity supporter | Google Venkat Scholarship winner | Speaker

Friday September 12, 2025 11:00am - 11:25am PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type OPEN Session
Tracks OPEN Session, DataWeek, Data Governance and Security (DataWeek), Data Strategy / Analytics & Business Intelligence (DataWeek), Virtual
In-Person/Virtual Virtual

11:30am PDT

[Virtual] OPEN Session (DataWeek): Rearchitecting Data Processing for Today’s Demands

Friday September 12, 2025 11:30am - 11:55am PDT

VIRTUAL DataWeek -- Main Stage

Rajan Goyal, DataPelago, CEO & Co-founder

IT leaders face mounting pressure to leverage their organization’s data for genAI and lakehouse analytics. Yet, with data volumes doubling every two years and nearly 90% of new data being unstructured, traditional data processing architectures can’t keep up. Existing systems were designed for structured data and CPU-based computing, and businesses are finding they struggle with processing latency, high costs, and siloed data. In fact, more than 80% of IT leaders say that data silos are hindering digital transformation.

To drive value in today’s data environment, IT leaders need to implement new data processing architectures that are designed to handle massive volumes of complex data and are capable of taking advantage of the accelerated hardware (GPUs, FPGAs, CPU/SIMD, etc.) available in today’s cloud environments.

In this session, DataPelago CEO Rajan Goyal will outline the shortcomings of current data processing architectures and introduce attendees to the Universal Data Processing Engine, a software solution that sits between a data lake and query engine. Overcoming the shortcomings of traditional processing models, the UDPE is designed to handle all types of data (structured, unstructured, semi-structured) and work on top of any hardware. Integrating seamlessly into any tech stack, it accelerates data processing speed by 2-3x while cutting processing costs by 30-60% — enabling organizations to utilize all of their data for lakehouse analytics and AI workloads.

Speakers

Rajan Goyal

CEO & Co-founder, DataPelago

Rajan is the co-founder and Chief Executive Officer of DataPelago, the company revolutionizing data processing for the accelerated computing era. His expertise and visionary approach have been instrumental in shaping the future of data infrastructure and processing. A seasoned innovator... Read More →

Friday September 12, 2025 11:30am - 11:55am PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type OPEN Session
Tracks DataWeek, OPEN Session, Data Tools / Technology and Management (DataWeek), Virtual
In-Person/Virtual Virtual

1:00pm PDT

[Virtual] PRO Session (API): FLEX APIs: Experience-Driven Design for an AI-Powered Future

Friday September 12, 2025 1:00pm - 1:25pm PDT

VIRTUAL API World -- Workshop Stage A (PRO)

Chris Hood, PolyAPI, CMO

APIs weren’t built for humans. But experiences are. In this keynote, Chris Hood challenges traditional API design models and introduces experience-based API design. This approach begins with user intent and adapts to how people and machines interact with systems today. Learn why naming conventions matter, how AI is changing who consumes APIs, and how the FLEX design model enables dynamic, on-demand APIs that respond to consumer intent and context. These adaptable interfaces do more than support the customer journey, they evolve with it.

Speakers

Chris Hood

CMO, PolyAPI

Chris Hood is a strategist, author, and keynote speaker who has become a pivotal thought leader in customer success, APIs, and digital strategy. With over 35 years of experience, he has helped organizations align technology with business goals. Chris built his first API platform in... Read More →

Friday September 12, 2025 1:00pm - 1:25pm PDT
VIRTUAL API World -- Workshop Stage A (PRO)

API World

Talk Type PRO Session
Tracks API World, APIs and AI (API World), API Design and Development (API World), Virtual
In-Person/Virtual Virtual

1:00pm PDT

[Virtual] PRO Session (DataWeek): Compute for Your AI Model: GPUs, LPUs, TPUs and Beyond..

Friday September 12, 2025 1:00pm - 1:25pm PDT

VIRTUAL DataWeek -- Main Stage

Kushaagra Goyal, Rubrik, Tech Lead

In the rapidly evolving landscape of computing, Graphics Processing Units (GPUs), Language Processing Units (LPUs), and Tensor Processing Units (TPUs) play pivotal roles in accelerating complex tasks, particularly in machine learning and artificial intelligence.

GPUs are renowned for their parallel processing capabilities, making them ideal for rendering graphics and handling large datasets. LPUs are specialized for optimizing natural language processing tasks, enhancing efficiency in understanding and generating human language. TPUs, developed by Google, are tailored specifically for training and inference of machine learning models, offering significant performance advantages for large-scale AI applications.

As we explore these technologies, we'll also look at emerging processing units designed for specific AI use-cases and the future of computational advancements.

Join me to dive into the intricacies of these processing units, their applications, and what lies ahead in the world of computing technology.

Speakers

Kushaagra Goyal

Tech Lead, Rubrik

Kushaagra Goyal is an accomplished technology leader with deep expertise in engineering and AI infrastructure. He holds a Bachelor’s degree from the Indian Institute of Technology, Delhi (2016), and a Master’s degree from Stanford University, where he developed a strong foundation... Read More →

Friday September 12, 2025 1:00pm - 1:25pm PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type PRO Session
Tracks DataWeek, AI Models and Management (DataWeek), Data Science and Machine Learning (DataWeek), Virtual
In-Person/Virtual Virtual

1:30pm PDT

[Virtual] PRO Session (API): Scalable Region Failover for Microservices: Platform-Level Resilience at Intuit

Friday September 12, 2025 1:30pm - 1:55pm PDT

VIRTUAL API World -- Workshop Stage A (PRO)

Chaitanya Bhatt, Intuit, Principal Engineer
Akshay Pratinav, Intuit, Staff Software Engineer

In today’s distributed architecture, where APIs are the backbone of product experiences, regional outages can break critical customer flows and severely impact business continuity. At Intuit, we’ve built a Kubernetes-native platform that delivers seamless, scalable region failover for hundreds of microservices—enabling consistent API resiliency across the board.

This talk explores how our platform empowers services to declaratively onboard and define key components for failover—compute, databases, and routing. During a failover, the platform automatically pre-scales infrastructure in the secondary region based on live traffic data, promotes standby databases (Amazon Aurora, Redis, and others), and orchestrates traffic redirection via AWS Route 53 or Istio DNS controls. This eliminates manual guesswork, reduces recovery time, and helps ensure that API endpoints remain responsive and stable—even in disaster scenarios.

We’ll also compare the risks of non-standard, ad hoc failover strategies—which often result in brittle systems, untested assumptions, and high operational overhead—with the advantages of a standardized, platform-level approach. Our framework simplifies validation, reduces testing complexity, and enables teams to run game days and chaos experiments with greater confidence and less stress.

Attendees will walk away with practical techniques and patterns for building resilient APIs in a microservices-first world, and learn how investing in a failover platform creates a stronger foundation for disaster preparedness, operational excellence, and a better end-user experience.

Speakers

Chaitanya Bhatt

Principal Engineer, Intuit

Chaitanya Bhatt, an engineering leader with 18 years of experience, specializes in Kubernetes and large-scale applications. As a Principal Engineer at Intuit, he focuses on infrastructure reliability and performance. Chaitanya has made significant contributions, including designing... Read More →

Akshay Pratinav

Staff Software Engineer, Intuit

Staff Software Engineer with extensive experience building distributed systems at scale. Passionate about system design, developer productivity, and leading high-impact engineering initiatives. Outside of work, I enjoy travelling and watching movies

Friday September 12, 2025 1:30pm - 1:55pm PDT
VIRTUAL API World -- Workshop Stage A (PRO)

API World

Talk Type PRO Session
Tracks Testing / Deploying & Monitoring APIs and Services (API World), API World, Virtual
In-Person/Virtual Virtual

1:30pm PDT

[Virtual] PRO Session (CloudX): Autoscaling in Cloud Native Kubernetes Environment

Friday September 12, 2025 1:30pm - 1:55pm PDT

VIRTUAL CloudX -- Main Stage

Geetha Anne, Pure Storage, Sr Manager, Customer Solutions Group

In this session, we’ll dive into the complexities of Kubernetes autoscaling, focusing on an essential aspect of distributed systems i.e, scaling. We'll begin by examining the built-in tools Kubernetes offers for each of these dimensions, and then move on to advanced open-source projects like KEDA, which provide greater flexibility and efficiency for various scaling scenarios.
Key takeaways for attendees include:
* Common challenges and solutions that enhance or replace traditional methods of scaling
* Strategies to evaluate application workloads and select the optimal tailored autoscaling strategy.
* Best practices to optimize deployments for dynamic workloads, ensuring peak performance while maintaining cost-efficiency.

Speakers

Geetha Anne

Sr Manager, Customer Solutions Group, Pure Storage

Geetha is a Solutions Architect and Customer Success professional specializing in big data management , storage , Kubernetes and Durable execution, with expertise in cloud-native and on-premises solutions. She ensures customer success and maturity by delivering effective, simplified... Read More →

Friday September 12, 2025 1:30pm - 1:55pm PDT
VIRTUAL CloudX -- Main Stage

CloudX

Talk Type PRO Session
Tracks CloudX, Cloud Infrastructure Design and Architecture (CloudX), Virtual
In-Person/Virtual Virtual

1:30pm PDT

[Virtual] PRO Session (DataWeek): Securing Multi-Tenant Data Warehouses: How Federated Learning Revolutionizes Privacy

Friday September 12, 2025 1:30pm - 1:55pm PDT

VIRTUAL DataWeek -- Main Stage

Jayant Tyagi, Salesforce, Lead Member of Technical Staff

Enterprises are being forced to reconsider how they manage sensitive data in cloud data warehouses due to data privacy laws and security incidents. In multi-tenant settings, traditional centralized analytics techniques are becoming more and more susceptible, putting businesses at risk for data leaks and regulatory issues.
In this session, we'll look at how federated learning is transforming data warehouse security while preserving analytical capabilities. Attendees will learn how businesses can use safe federated ways that preserve sensitive data while still obtaining insightful information, based on current research in privacy-preserving technologies.

Speakers

Jayant Tyagi

Lead Member of Technical Staff, Salesforce

Jayant Tyagi is a seasoned full-stack engineer with 13 years of experience in designing and developing high-scale applications. As a Lead Member of Technical Staff at Salesforce, he has played a pivotal role in building and optimizing enterprise applications, spearheading innovations... Read More →

Friday September 12, 2025 1:30pm - 1:55pm PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type PRO Session
Tracks DataWeek, Data Warehousing and Storage (DataWeek), Virtual
In-Person/Virtual Virtual

2:00pm PDT

[Virtual] OPEN Session (DataWeek): Transforming Seller Onboarding in Retail: Responsible AI, RAG, and Risk Management

Friday September 12, 2025 2:00pm - 2:25pm PDT

VIRTUAL DataWeek -- Main Stage

Banani Mohapatra, Walmart, Senior Manager, Data Science
Bhavnish Walia, Amazon, Senior Risk Manager AI/ML

Onboarding new sellers onto retail platforms like Walmart and Amazon involves a complex, multi-step process designed to mitigate fraud and ensure compliance with global regulations. One of the most critical and cumbersome steps is Know Your Customer (KYC) verification, requiring sellers to upload documentation for identity verification, business registration, and compliance checks. This manual review process often leads to long approval times and delays, frustrating legitimate sellers and creating operational bottlenecks for compliance teams.
To address these challenges, we leveraged foundational models with custom prompting strategies, in-document summarization, and retrieval-augmented generation (RAG) to ground responses in trusted data sources, powered by open-source LLM APIs. By automating document analysis and augmenting human reviewers with AI outputs, we reduced overall onboarding time by more than 20 percent, improving seller experience and operational efficiency.
However, deploying AI into a regulated process like KYC required a robust responsible AI framework combining scalability with governance. We implemented guardrail models to flag edge cases and ensure human oversight, enforced strict data anonymization protocols to protect sensitive information, and applied privacy-preserving techniques for model training. We also established a rigorous validation pipeline to test outputs against regulatory standards, mitigating risks such as hallucinations and interpretability gaps.
This talk offers actionable insights for data scientists, compliance officers, regulators, and machine learning practitioners working at the intersection of AI, risk management, and regulatory compliance. Presented by Bhavnish Walia, Senior Risk Manager at Amazon, and Banani Mohapatra, Senior Data Science Manager at Walmart, attendees will walk away with a practical framework for deploying AI in sensitive domains—covering risk management strategies, scalable AI architectures aligned with compliance, and key lessons on balancing innovation with accountability.

Speakers

Bhavnish Walia

Senior Risk Manager AI/ML, Amazon

Bhavnish Walia is a Senior Risk Manager at Amazon, where he leads AI Risk Management efforts focused on developing large language model (LLM) frameworks for data governance and regulatory compliance. He ensures the safe and compliant deployment of AI systems at scale. With over 12... Read More →

Banani Mohapatra

Senior Manager, Data Science, Walmart

Banani Mohapatra is a data science leader with 12+ years of experience in e-commerce, payments, and real estate, specializing in machine learning, generative AI, LLMs, and causal AI. She leads a global data science team at Walmart, driving subscription growth with multi-billion-dollar... Read More →

Friday September 12, 2025 2:00pm - 2:25pm PDT
VIRTUAL DataWeek -- Main Stage

DataWeek

Talk Type OPEN Session
Tracks DataWeek, OPEN Session, Data Governance and Security (DataWeek), Virtual
In-Person/Virtual Virtual