AI Data Builder

Mekari; Jakarta Pusat; On-site; IDR 180,000,000 - 300,000,000; Full time

Job summary

Mekari in Jakarta seeks an AI/Data Builder to own AI/ML, LLM, data products, and end-to-end features. You will ensure probabilistic behavior quality, data quality, model behavior, and guardrails while shipping useful, safe AI capabilities.

Stay close to users, ship and measure impact, collaborate with Platform and Product Builders, and continuously improve prompts, models, and data pipelines based on real usage.

Qualifications

• 2-3+ years in full-stack engineering or AI/data product development.
• Experience building AI/data-powered features and services.
• Ability to design evaluation datasets and guardrails.
• Knowledge of grounding, hallucination control, and trust.
• Strong collaboration with product and platform teams.

Responsibilities

• Define and ship AI/data capabilities that are useful, safe, and measurable.
• Stay close to user workflows to set 'good' vs. 'unacceptable' AI behavior.
• Build AI services, ML models, LLM workflows, and data pipelines.
• Configure prompts, retrieval logic, and model settings.
• Validate quality with eval sets, golden datasets, and human review.
• Manage risk: latency, cost, privacy, drift.

Skills

AI/ML development

LLM & RAG

Data pipelines

Model evaluation

Guardrails & safety

Prototyping & shipping

User-centered design

Tools

Python

TensorFlow/PyTorch

SQL/BigQuery

Jupyter

Job description

Mekari is Indonesia's no. 1 Software-as-a-Service (SaaS) company. With our ecosystem of software solutions—including Mekari Jurnal, Mekari Talenta, Mekari Qontak, and Mekari Flex, we aim to facilitate entrepreneurs and leaders as they accelerate the digital transformation of their businesses.

In our 10+ years of journey we have reached over 3 Million platform users, and we're not planning to stop any time soon. We need more people like you: builders and owners with calculated ambition who are ready to #ElevateThroughImpact and raise Indonesia's software standard.

Job Description:

The AI/Data Builder is an AI-native Builder who owns AI, ML, LLM, data, and non-deterministic system quality. This role expects to stay close to users and ships end-to-end, user-accessible AI/data products, features, modules, workflows, services, and decision-support capabilities. What sets this role apart is deeper accountability for the quality of probabilistic behavior, data quality, model behavior, evaluation systems, AI trust, guardrails, and continuous improvement. The mission: build, ship, measure, and improve AI/data capabilities that are useful, safe, trusted, measurable, and valuable for users.

• User & product alignment. Stay close to workflows, feedback, and production behavior to define what "good vs. unacceptable" AI behavior looks like for the use case.
• Build & ship.Develop AI services, ML models, LLM workflows, RAG systems, data pipelines, and model integrations as real product capabilities (not just experiments).
• Configure & construct. Own prompts, retrieval logic, model configs, orchestration flows, and data transformations.
• Validate quality. Test behavior through eval sets, golden datasets, scenario/adversarial tests, regression evals, and human review.
• Manage risk. Handle grounding, hallucination, bias, privacy, latency, cost, and drift concerns.
• Release responsibly. Ship with guardrails, fallbacks, monitoring, and human-in-the-loop controls.
• Iterate. Continuously improve prompts, models, and pipelines based on real usage data.
• Cross-functional collaboration. Partner with Platform Builders (deployment, agent tuning) and Product Builders (user value, trust boundaries, release readiness).

Requirements:

• 2 - 3+ years in full-stack engineering or applied software development using AI/data use cases.
• Experience with AI, ML, LLM, RAG, data, analytics, and intelligent product systems.
• Proficiency in prompt, context, retrieval, model, orchestration, and data-system design.
• Experience with data pipelines, data quality, freshness, lineage, completeness, and correctness.
• Demonstrated ability with evaluation design, golden datasets, regression evals, scenario tests, and human review.
• Working knowledge of grounding, hallucination control, refusal behavior, guardrails, safety, and trust.
• Experience monitoring for drift, quality degradation, latency, cost, user feedback, and production behavior.
• Track record shipping AI/data-powered products, features, modules, workflows, services, or decision-support capabilities.
• Preferred Traits
  • Experience collaborating with Platform Builders on agentic build workflows, especially agent tuning, evaluation, and improvement.
  • Comfort defining risk tolerances and behavior standards for non-deterministic systems (i.e., systems where output varies based on models, prompts, retrieval, data, context, or user input, and can't be verified by exact-output tests alone).
• Mindset
  • A builder who takes deep ownership of AI/data behavior quality — someone who treats evaluation, monitoring, and continuous improvement as core parts of shipping, not afterthoughts, and who partners naturally with Product and Platform counterparts (this role isn't research-only, model-only, or analytics-only — it's about shipping real AI/data-powered capability and managing its quality in production).

We wish you the best. Hope to see you around soon!