Fraud Operations Analyst

Nelo is a leading consumer fintech and e-commerce platform in Mexico, with >$500MM in annualized GMV and >$75MM in annualized revenue.

Our mission is to increase the buying power of consumers in Latin America, and we are doing so by building a modern alternative to credit cards. Nelo has raised over $40M of venture capital from investors including Homebrew, Two Sigma Ventures and Susa Ventures. Nelo has additionally raised a $100M asset credit facility from Victory Park Capital.

Our lean team includes experienced leaders from top technology companies including Uber, Amazon, Rappi, and DiDi.

We pride ourselves on our velocity, intellectual rigor, and efficiency. Nelo has offices in Mexico City and New York City.

Fraud at Nelo is moving toward an agent-operated model. AI does most of the volume. Humans set the standard. This role is the operating layer underneath that strategy, and it owns two things end-to-end: the AI fraud investigation stack, and AML.

On the AI side, you are the person who decides what good signal looks like, curates the seed cases that train the agents, writes the instructions that tell them where to look and what to ignore, and designs the JSON schemas they read from and write into. You build the filter that sits between CX intake and the agents so we feed them signal instead of noise. When new fraud MOs emerge, you catch them and feed the agents before performance decays.

On the AML side, you own casework, SAR narratives, escalations, and the operating cadence that keeps Nelo defensible if and when a regulator walks in. This is low-frequency, high-stakes work. It cannot be a side task and we are not treating it as one.

You will work closely with CX (the source of most raw signal), Risk, Legal & Compliance, and Engineering.

• Own the AI fraud agent stack: curate seed cases, refine instructions, design the JSON schemas, run the feedback loop that sharpens them over time.

• Build and maintain the filter between CX intake and the agents. Define what "good signal" looks like and codify it.

• Investigate transactional fraud and identity theft cases end-to-end, especially the ones the agents flag for human review.

• Stay current on local fraud MOs (SIM swap on Telcel, OXXO cash-outs, account takeovers, synthetic identity) and feed that intelligence into both the agents and the broader Risk function.

• Own AML operations: casework, escalations, SAR narratives, and the documentation that holds up under regulatory scrutiny.

• Write the playbooks, case taxonomies, and filter rules that let this function scale beyond you.

• You have done transactional fraud investigations on a card or BNPL portfolio and know what a chargeback-driven case actually looks like. You can read a dispute, pull the right signals, and reach a conclusion the issuer or network will accept.

• You have worked AML cases, written SAR narratives, and understand how transactional patterns and identity signals interlace. You know the regulatory bar and you treat it seriously.

• You are an operator, not just an investigator. You see a noisy intake pipeline and your first instinct is to design the filter, not to muscle through every ticket.

• You are comfortable with AI tools as part of your daily workflow. You can write a clear instruction for an agent, read what came back, and know whether it was wrong because of bad instructions, bad data, or a bad case to begin with.

• You write well. Tight, structured, unambiguous. Investigations, SAR narratives, agent instructions, and case taxonomies all live or die by the quality of the writing.

• You have close working instincts with CX. You understand they are the source of the raw signal and you treat the relationship that way.

• You want a queue to clear. The job is to design how the queue gets handled, not to be the queue. If you are happiest closing tickets, this will frustrate you fast.

• You see AML as paperwork. The cost of an AML failure at a company like Nelo is not absorbable. If you treat compliance as a checkbox, you will not enjoy the level of rigor expected here.

• You think AI agents are a gimmick or, on the other end, that they run themselves. Both views break this role. The agents need an operator and a critic, every day.

• You need polished systems handed to you. Some of this exists. Most of it does not. You are building the operating model, not inheriting one.

• You are not based in CDMX or not willing to relocate. Most intake is in Spanish, MOs are local, and the CX team you depend on works from the office.

• ~60 people across CDMX and New York. Lean, fast, opinionated about quality.

• This role is based in Mexico City and is expected to be in-office given the proximity required with CX and Risk.

• Every team at Nelo uses AI in daily workflows. Fraud is no exception.

• Internal tooling: Slack, Linear, Metabase, Kustomer on the CX side, and an in-house case management layer for fraud that you will help shape.

• Demonstrated ownership of transactional fraud investigations (credit card, BNPL, or comparable consumer credit product).

• Identity theft casework and understanding of how it intersects with transactional fraud.

• AML experience, including SAR narrative writing and familiarity with the regulatory expectations of a consumer financial institution.

• Native Spanish, professional English.

• Based in Mexico City or willing to relocate.

Nice to have:

• Practical fluency with AI tools applied to investigations or operations work.

• CFE, CAMS, or comparable certification.

• Intro call

• Take Home Exercise

• On-site Interview

• Background and references check

• Offer

• Competitive salary and equity

• 100% medical, dental, and vision coverage

• Unlimited PTO

• Extended parental leave