Business Analyst

mscope; Madrid; Híbrido; EUR 42.000 - 65.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

25 days of vacation (and your birthday

Hybrid mode: 2 days in office 3 days

Descripción de la vacante

mscope in Madrid is seeking a Business Analyst to help build the scoring methodology and taxonomy used to rate tens of thousands of companies across multiple countries. You will join the Business Analysts, product squads and report to the CPO from day one.

You’ll work with data engineering and data science, define thresholds, validate changes before release, and explain model evolution to internal teams and clients. A 3+ year track in analytics and daily use of LLMs are required.

Formación

• 3+ years as a business, financial, or data analyst in fintech, banking, or market research.
• Ability to define rules, edge cases and document them clearly.
• Experience evaluating AI agents and prompt engineering.
• Excellent attention to detail and structured thinking.

Responsabilidades

• Build the methodology behind our scoring, classification and taxonomy together with the squads. Propose eligibility rules, thresholds and metric definitions, document them with examples and versioning, and validate each change before release.
• Review PRDs and designs from a business and financial-statements angle and give engineering written answers, so that nobody builds ahead of the definition.
• Analyse the sector and macro context by country (growth, cycles, regulation) and bring that reading into model discussions on weights, thresholds and taxonomy changes.
• Prepare and read product and data-quality metrics (classification confidence, accuracy, adoption in Mixpanel) for the metric review and turn them into recommendations for PMs.
• Validate the data that drives client-facing outputs, meaning financials, ratings and microsector assignments, before each release.
• Use 𝗟𝗟𝗠𝘀 everyday as a working tool. Analyse data and documents with Claude or ChatGPT, prototype rules and classifiers with them, evaluate the output of our own AI agents and write the evals that tell us whether they are right.
• Help explain the model and its evolution to internal teams and, together with product, to clients and partners.

Conocimientos

LLM usage

Prompt engineering

Structured thinking

Cross-functional collaboration

Attention to detail

AI evaluation

Herramientas

Claude Code

Descripción del empleo

The role

Our product rates more than 𝟰𝟬𝟬,𝟬𝟬𝟬 companies across 𝟳 countries using a proprietary scoring model and a three-level sector taxonomy. That methodology is built and maintained jointly by the Business Analysts and the product squads, and you will join that work from day one. You will help decide how a company qualifies for a metric, what threshold separates a signal from noise, how a microsector is defined and documented, and whether a change to the model is ready to reach clients. You will work with another Business Analyst, the four product managers, the product designer, data engineering and data science, and you will report to the CPO.

What you’ll do

• Build the methodology behind our scoring, classification and taxonomy together with the squads. Propose eligibility rules, thresholds and metric definitions, document them with examples and versioning, and validate each change before release.
• Review PRDs and designs from a business and financial-statements angle and give engineering written answers, so that nobody builds ahead of the definition.
• Analyse the sector and macro context by country (growth, cycles, regulation) and bring that reading into model discussions on weights, thresholds and taxonomy changes.
• Prepare and read product and data-quality metrics (classification confidence, accuracy, adoption in Mixpanel) for the metric review and turn them into recommendations for PMs.
• Validate the data that drives client-facing outputs, meaning financials, ratings and microsector assignments, before each release.
• Use 𝗟𝗟𝗠𝘀 every day as a working tool. Analyse data and documents with Claude or ChatGPT, prototype rules and classifiers with them, evaluate the output of our own AI agents (classification, enrichment, assistant) and write the evals that tell us whether they are right.
• Help explain the model and its evolution to internal teams and, together with product, to clients and partners.

What we're looking for

• 𝟯+ 𝘆𝗲𝗮𝗿𝘀 as a business analyst, financial analyst or data analyst, ideally in fintech, banking, private equity, private credit or market research.
• Structured thinker who breaks an ambiguous question into definitions, rules and edge cases before proposing an answer, and writes it down so others can build on it.
• Advanced, daily use of 𝗟𝗟𝗠𝘀 (Claude, ChatGPT, Gemini or similar). You know how to prompt them well, how to structure a task so the output is reliable, how to check them when they are wrong and where they should not be trusted. Experience evaluating or building AI agents, or working with tools like 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲.
• High attention to detail, since your definitions drive outputs that clients see.
• Comfortable working across several squads at once and managing competing requests.

What we offer

• 25 days of vacation (and your birthday off!)
• Hybrid mode: 2 days in office 3 days in remote
• If interested, please email [e-mail]