About TRMS

The Transaction Risk Management (TRMS) team at Amazon builds platforms to ensure that every eCommerce transaction is trusted, safe and protected from fraud. Browse through the list of groups below that are a part of this dynamic team.

  • Abuse Prevention Services
  • Bangalore Platform Services
  • Buyer Feedback
  • Buyer Fraud Investigations
  • Buyer Fraud Detection
  • Communication Manager
  • Compliance Program
  • Customer Account Protection
  • Dispute Management
  • Investigations and Information
  • Machine Learning
  • Merchant Quality
  • Payments Fraud Management
  • Risk Analytics
  • Risk Platform Services
  • Seller Performance
  • Support Engineering

Abuse Prevention Services

Abuse Prevention Services strives to protect Amazon businesses exposed to customer abuse while maintaining the highest level of customer experience for our good customers. This means building highly sophisticated, data-centric systems that can detect abusive patterns across millions of transactions. We build highly scalable, flexible and distributed systems that utilize the power of data at every step - compute predictive variables, build models using machine learning algorithms and plug into different pipelines to prevent abusive transactions from taking place. As Amazon businesses grow, abusers morph to find new ways to take undue advantage, our engineers and scientists are constantly innovating to stay ahead of the game and protect Amazon and our customers.

Bangalore Platform Services

Bangalore India

TRMS Bangalore Risk Platform team builds common solutions for all of TRMS. We provide the rules engine that every TRMS teams use to build their sophisticated models to detect fraud patterns for all buyers, sellers, digital, and AWS businesses and eliminate risks. Software engineers in this team excel at scaling technologies to highest levels. We enable an UI portal for Amazon customers where they can recover their accounts in a self-service fashion from suspended or compromised states.

Buyer Feedback

Let's say you want to buy something on Amazon. You visit the product detail page and find 10 different third-party Sellers selling the same item using the Amazon Marketplace platform. How do you decide which Seller to buy from? Price would definitely be an important consideration. But let's say most sellers are selling around the same price. Whom to pick? One of the most important factors that can help or influence your decision is how good a particular Seller is. But how do you know unless you actually buy from the Seller? Amazon solves this problem for you currently by gathering feedback from other Buyers who have bought from a particular Seller and presenting it in the form of a Seller Rating (1 – 5 stars), feedback break-up (positive, neutral, negative) and actual feedback text. If the feedback for a particular Seller is much better than another Seller, you might even be ok to pay a bit of a price premium in order to buy from the better Seller and such is the nature and importance of feedback. Its very important for Buyers as it helps them a lot with their buying decision. Its very important for Sellers as it helps them understand what are they doing right and what can they improve upon. Its very important for Amazon as fueling better buying decisions for Buyers and increasing their trust in Sellers means more transactions.

The Seller Feedback Team owns the complete feedback program for Amazon end-to-end – right from the customer experience to all the engineering systems. We own how to encourage Buyers to leave feedback, how to process the feedback data, how to show the feedback to other Buyers, how to the show the feedback data to the Seller, how to handle responses from Sellers on their feedbacks, how to resolve conflicts – basically all things related to the feedback program. And with the surge in social networking and information sharing and collaboration, there is a lot of scope to come up with new and innovative ideas to solve these problems better and better. As part of the feedback team, you would get exposure to both Amazon Buyer and Amazon Seller customer experience, large–scale distributed systems that crunch the data, computer science algorithms that work in the real world, and an opportunity to work on and deliver lot of impactful things that fuel the Amazon business. Come – work hard, have fun and make history – as part of this wonderful and enriching journey!

Buyer Fraud Investigations

Checking all customer orders for possible fraud requires skilled investigators to review and examine suspicious activity. To be successful this worldwide investigations team uses powerful and intuitive software. This software has to retrieve, organize and display data from numerous Amazon services with accuracy, speed and without bias for a fair investigation. Software developers in this team design both large-scale distributed computing systems and yet also have a passion for developing simple and effective user interfaces providing for a very valuable set of blended skills.

Buyer Fraud Detection

Amazon is constantly under attack by sophisticated criminals from all over the world. Armed with stolen credit cards -- and often, hacked customer accounts -- they place fraudulent orders in hopes of getting products for free. The Buyer Fraud Management team is charged with pioneering new techniques to detect and neutralize the fraudsters without impacting our legitimate customers. Our software scans every order at Amazon, computing hundreds of individual risk indicators per order and leveraging sophisticated machine learning algorithms to identify the orders that need to be stopped. We build highly scalable and fault-tolerant distributed systems to handle Amazon's massive traffic volume. We must also make super-fast decisions to accommodate Amazon's new digital download businesses, such as Kindle and MP3. Fraudsters are continuously getting better at disguising themselves, so every engineer is encouraged to innovate and contribute towards our long-term vision.

Communication Manager

The Communication Manager team is responsible for building and maintaining the global messaging service used to manage communication between buyers and sellers in Amazon’s marketplace. The service ensures email privacy for both buyers and sellers, and screens every email sent through Amazon systems. Our system supports trusted communications by safeguarding customers from being targeted by unscrupulous people with unwanted marketing, spam, phishing and fraud scams outside Amazon. Software engineers on this team take pride in building highly-scalable systems that ensure millions of messages are delivered reliably to Amazon customers.

Compliance Program

Bangalore India

Amazon's AML Compliance Program protects all Amazon businesses and Subsidiaries from aiding or associating with money laundering and terrorist financing activities. This program ensures that the identifying information a customer provides is accurate and legitimate. It enforces that the customer does not belong to any of related government sanctions lists of individuals or countries including terrorist organizations, and monitors all transactions activities for all businesses for any suspicious money laundering activity. We build systems that can scale for every customer registrations, all transactions including every order and shipment on Amazon, every payment change and every address change. We integrate with external vendors across the globe to access customer databases maintained by government organizations worldwide.

Customer Account Protection

Making Amazon the safest and most trusted place to shop online means protecting our buyers, sellers, and developers against identity theft. The CAP team at Amazon works to stop phishing attacks, malicious software, and automated attacks from hurting our customer's experience on all Amazon sites. This team is a highly flexible and motivated group that operates autonomously to protect our customers utilizing the latest machine learning algorithms and large scale software analysis tools. This team is on the "front line" in the fight against identity theft.

Dispute Management

Dispute management team provides global storage for different kinds of disputes and currently stores dispute related to buyers Atoz guarantee claims and chargeback for Amazon marketplace business. This team also built and support components on buyer and seller facing website which enables submission of disputes, re-presentation of disputes or providing refund for a dispute. It also provides components that are leveraged by risk investigators and customer service team to answer buyer/seller questions. Analyst team leverages dispute data to close loop on fraudsters. Systems build in this space has huge impact on companies bottom line and you can measure how much $$ is added to FCF (free cash flow) for every line of code.

Investigations and Information

Providing excellent service to our customers when they're the victims of identity theft requires a personal touch. At Amazon TRMS, skilled investigators resolve on customer concerns and investigate suspicious activity. Supporting this worldwide investigation team with powerful and intuitive tools is quite a challenge. Not only do these tools need to research and diagnose issues from across the whole of Amazon and subsidiaries, they must provide this data quickly and without bias for a fair investigation. Software engineers in this team regularly pull double duty being both large-scale distributed computing designers, as well as HCI experts, a very useful blend of skills.

Machine Learning

Bangalore India

A dedicated team formed to develop Machine Learning services and applications which can be leveraged by Seller Services and other internal Amazon customers such as Customer Service, Amazon Local and Amazon Retail. The team will build new business solutions which leverage machine learning, such as those for intelligent seller support and customer service, seller recruiting and driving sales growth. The team is expected to grow significantly based on its demonstrated wins and help architect a variety of successful machine learning applications across seller services and retail.

Merchant Quality

A marketplace thrives with diverse and trusted sellers. Sellers thrive on a safe marketplace. The Merchant Quality team is responsible for protecting good sellers and identifying fraudulent sellers on Amazon marketplace. This team provides advanced fraud detection mechanisms for all seller types – Selling On Amazon, Check Out By Amazon, Webstore, Product Ads, AWS (EC2), Kindle Digital Publishers and Amazon Associates. These fraud detection mechanisms include multi-level checks during seller registration and post registration catalog activities by doing out-of-band authentication, applying advanced machine learning techniques to analyze data. The Merchant Quality Team is also tasked with building a new platform to prevent fraudsters from re-infiltrating our services. Our engineers are constantly innovating, exploring new technologies to scale our software and support new businesses, and discovering new ways to strengthen our services to protect Amazon and its customers.

Payments Fraud Management

Bangalore India

Amazon Payments TRUST team protects FPS customers and merchants from financial fraud. This means building scalable systems which can detect fraud in milliseconds and enable fast and secure money transfer between individuals. We build sophisticated risk models that can detect fraud patterns across millions of transactions for businesses outside of Amazon like online games and crowd funding.

Risk Analytics

The team analyzes large transactional data, history data and stream data for the knowledge of business development and problem solving. Modern machine learning and statistics skills are widely used to build predictive models and algorithms for fraud detection and credit risk management. The team consists of PHD and MS from different quantitative areas and works together with engineers and operations to development end-to-end strategies and real time implementation and monitoring. Innovation and simplification are key to scale up with the tremendous growth of Amazon business. Research leading to state of art models and algorithms to better solve business problems are highly encouraged.

Risk Platform Services

Transaction Risk Management at Amazon is a big job. Maintaining a clean and robust software architecture provides stability and scalability for our future. The Risk Platform Services team at Amazon works to provide solutions to common problems such as analytics, machine learning, high availability data storage, caching, and more. The software architectures and designs created here are used throughout TRMS and within other teams at Amazon. Software engineers in this team excel at scaling technologies to incredible levels.

Seller Performance

The Seller Performance Team manages customer experience provided by third-party merchants selling on the Amazon Marketplace platform. We own the systems, processes, algorithms and software used to automatically identify the best and underperforming merchants on the platform. Our goal is to help merchants provide superb customer service so that they can grow to their fullest potential. We are relentless in our pursuit of raising the bar with respect to the customer experience provided. Software engineers in this team excel at scaling technologies and handling large volumes of data for analysis.

Support Engineering

Our team provides systems support to internal TRMS engineering teams as they build business-critical fraud detection platforms. We maximize the efficiency of TRMS software development resources by reducing their operational overhead. We bring department level economies of scale to deliver team level projects within organization-wide infrastructure initiatives. We accomplish this by maintaining specialization in common engineering tasks, by developing and maintaining efficiency matrices, and by documenting and executing standard operating procedures.