Lead Data Analyst

LEAD ANALYST, DATA

Vaudit | Bangkok

Location: Bangkok, onsite

Remote start: Possible, with relocation to Bangkok required

Reports to: COO

Team: Data

Scope: Own the data and analytics function, including the analysts who deliver it THE ROLE

Every invoice our clients pay was calculated by the vendor who measured what was delivered. Virtually no one verifies their underlying math. We built the platform that does.

Your primary objective is to uncover the patterns hidden within that billing data. Our detection logic exists because someone thoroughly analyzed the data, identified a systemic issue, and proved it was reproducible. Uncaptured violation types currently exist in our datasets; your role is to discover them, test whether cross-category patterns hold, and convert those insights into production-ready detection logic.

The second half of this role centers on core analytics—acting as the direct bridge between raw data and commercial strategy. You will define primary metrics, investigate performance shifts, and evaluate financial opportunities before engineering resources are committed. This is not a reporting function; it is about delivering actionable answers.

This is a player-coach role, heavily weighted toward hands-on execution. You will spend roughly two-thirds of your time working directly in the data, with the remainder dedicated to setting analytical standards, reviewing work, and developing the team. If you are seeking a purely managerial or hands-off position, this role is not the right fit.

WHAT THIS ROLE IS NOT

• Not deep model research. No deep learning papers, no novel architectures, no publishing. We need working detection logic, not a PhD thesis. If your instinct is to reach for the simplest method that answers the question, you'll do well here.
• Not a report factory. Ad hoc chart requests are not the job, and we'll back you when you push them away.

We are a startup, so the boundaries are not clean. If a pipeline is broken and it is blocking your analysis, we expect you to go fix it rather than file a ticket and wait.

CORE CAPABILITIES

Technical depth

Advanced SQL: window functions, CTEs, incremental models, and awareness of what a query costs on a very large table. Python for analysis, plus regression, forecasting, cohort analysis, and anomaly detection. Writes notebooks someone else can rerun without you in the room.

Discovery

Given a vendor's raw billing export, finds the charges that do not reconcile against what was delivered, quantifies how often it happens, and shows it is systematic rather than a one-off. Checks whether the same pattern holds for a different vendor before generalizing.

Building what you find

Turns a validated pattern into detection logic with defined thresholds, a false positive rate you can state out loud, and documented edge cases. Ships it with engineering rather than handing over a notebook.

Leverage from AI

We are an AI-native company and we expect you to work like one. Uses AI to triage volumes of data no person could read and to build things that would not otherwise justify the effort. We care about the output, not the tool list.

Leading the team

Reviews analysts' work and sends it back when the method is wrong, not just when the number is. Sets the queue against business need rather than whoever asked loudest. Writes findings as short arguments and holds a position when someone senior pushes back.

Startup experience is required. You have worked somewhere without established process, clean data, or a function to hand problems to, and you were effective anyway.

STRONG PLUS

• Experience in a domain where the data itself is the product (audit, fraud, billing, payments, risk, adtech)
• Has worked with vendor or third-party data of variable quality
• Data engineering ability: pipelines, dbt or equivalent, orchestration. Not something you will own, but at our size the person who can unblock themselves is worth a lot.

We do not screen on degrees. Show us what you have built and what changed because of it.

TOOLING

BigQuery, Python, dbt, Metabase, Git

BENEFITS

Health insurance. We sponsor visas and work permits for international hires.

PROCESS

We start with a short call to assess fit against the criteria above, based on your past experience. We will also cover your relocation timeline. If it is a fit, we will walk you through the remaining steps from there.

We move quickly for the right person.

ABOUT VAUDIT

Vaudit is an AI-native agentic spend auditing platform built for enterprise. Our fleet of specialized AI agents audits spend across every major category: AI infrastructure (OpenAI, Anthropic, AWS Bedrock, Azure OpenAI, Google AI, Cohere, and every major model provider), cloud (AWS, GCP, Azure), advertising platforms, shipping and logistics carriers, SaaS, and operational vendors. The agents verify charges, detect billing violations, enforce spend guardrails, and recover money 24/7.

The underlying problem is the same in every category: the vendor sending the invoice is also the only party measuring what was delivered. We have audited over $1 billion in spend and recovered more than $50M for our clients. We are pre-Series A, SOC 2 Type 2 and ISO 27001 certified.