Data Scientist - AI (Collection)

Kredivo; Jakarta Pusat; On-site; IDR 240,000,000 - 360,000,000; Full time

Job summary

Kredivo in Jakarta seeks a Data Scientist – AI (Collection) to design, ship and measure AI solutions across the collection journey. You will build LLM-powered assistants, automate desk collection analysis, and develop predictive models with rigorous experiments.

Collaborating with Collection Ops, Product and Engineering, you will push AI transformations from early reminders to late-stage recovery, applying RAG, prompt engineering, and MLOps practices to production systems.

Qualifications

• Bachelor's or Master's in a quantitative field.
• 2+ years in data science or ML, preferably fintech/collections.
• Strong Python and SQL; hands-on with ML libraries.
• Hands-on with LLMs: prompt engineering, RAG, agents and APIs.
• Exposure to conversational AI or voice tech is a plus.
• Experience with experimentation and statistics (A/B testing).
• Production deployment or MLOps exposure is a plus.
• Fluent in Bahasa Indonesia and good English.
• Product-minded and hands-on: shipping, measuring and iterating.

Responsibilities

• Design and build LLM-powered collection assistants (WhatsApp chatbots).
• Audit and analyze desk collection calls/chats using STT and LLMs.
• Develop predictive models (propensity-to-pay, contactability).
• Design and analyze A/B tests to measure KPI impact.
• Collaborate with Engineering and Ops to deploy and scale solutions.

Skills

Python

SQL

ML libs

LLMs

A/B testing

MLOps

Bahasa Indonesia

English

Education

CS/Math/Engineering degree

Tools

scikit-learn

XGBoost/LightGBM

PyTorch

Airflow

Docker

Job description

Kredivo is hiring a Data Scientist – AI (Collection) in our Jakarta office to build AI-powered products that change how we collect payments from customers. This position will report to the Lead Data Scientist – Collection & SPG and will be accountable for designing, shipping and measuring AI solutions across the collection journey, from early reminders to late-stage recovery.

In this role you will also be responsible for driving AI transformation within Collection: building conversational AI (chatbots and voice bots/robocalls), automating desk collection audit processes, and applying LLMs to collection workflows. In addition, you will be responsible for proving business impact through rigorous experiments and working closely with Collection Ops, Product and Engineering to put models into production.

About the job:

• To design and build LLM-powered collection assistants (WhatsApp/in-app chatbots, AI voice agents/robocalls)
• To build automated audit and analysis of desk collection calls and chats, using speech-to-text and LLMs to check them against collection SOPs at scale and turn them into insights for Collection Ops
• To develop and improve predictive models (e.g. propensity-to-pay, contactability, best channel/time to contact) that decide who to contact, when and how
• To design and analyze A/B tests that measure impact on collection KPIs (roll rate, recovery rate, cost per collection)
• To work with Engineering and Ops to deploy, scale and iterate on solutions, and to make sure they follow our internal collection guidelines

About you:

• Bachelor's or Master's in Computer Science, Statistics, Mathematics, Engineering or a related quantitative field
• 2+ years of experience in data science or ML, ideally in fintech, lending, collections or customer operations
• Strong Python and SQL; hands-on with ML libraries (scikit-learn, XGBoost/LightGBM, PyTorch or similar)
• Hands-on experience building with LLMs: prompt engineering, RAG, function calling/agents, and LLM APIs or open-source models
• Exposure to conversational AI or voice tech (chatbot frameworks, STT/TTS, voice bot platforms) is a strong plus
• Solid grounding in experimentation and statistics (A/B testing, causal inference basics)
• Experience deploying models to production or working with MLOps tools (Airflow, Docker, cloud platforms) is a plus
• Fluent in Bahasa Indonesia and good English; able to explain technical work to non-technical stakeholders
• Product-minded and hands-on: you like shipping, measuring and iterating quickly

#LI-AL1