ML Engineer – Recommendation & Decision Engine

About SwanSat Group

SwanSat Group stands at the intersection of deep technology, precision intelligence, and social impact, uniting three powerful verticals to transform how nations, businesses, and communities make decisions.

At our core is SwanSat, a leader in satellite and geospatial intelligence. We harness cutting-edge SAR, optical, and thermal data, enhanced with advanced AI/ML algorithms, to deliver unmatched insights for agriculture, disaster management, defense, and climate resilience.

Through KhetPe, our AgTech platform, we place accurate crop intelligence directly into farmers hands, covering hyperlocal weather, irrigation advisory, yield prediction, and market linkage. Through BhoomiSure, our InsurTech platform, we enable automated, satellite-driven crop and disaster insurance with fast, dispute-free claim settlement.

Together, SwanSat Group builds mission-critical systems where reliability, explainability, and correctness directly impact livelihoods.

Role Summary

Own how signals are converted into safe, prioritised, and explainable actions across the platform.

Responsibilities

Decision System Design

• Design hybrid rule-based and ML-driven decision systems.
• Encode domain constraints and safety rules explicitly.

Ranking & Confidence

• Implement prioritisation and confidence scoring logic.
• Prevent misleading or over-confident recommendations.

Feedback & Learning

• Capture outcomes, confirmations, and reversals.
• Use feedback to improve decision quality over time.

Evaluation & Impact

• Measure uplift, error rates, and unintended consequences.
• Monitor and mitigate harmful or low-trust outcomes.

Tools & Technologies

• Python-based ML pipelines
• Backend APIs for scoring and recommendations
• Experimentation and logging frameworks

Why Join Us

Most technology roles optimise primarily for scale, speed, or convenience.

At SwanSat Group, those goals matter but they are balanced with a deep emphasis on correctness, reliability, and trust in real-world, high-stakes environments.

You will build systems that operate where assumptions break unreliable connectivity, incomplete data, real economic risk, and users who depend on predictable behaviour and clear outcomes. The software you ship influences real decisions made by farmers, insurers, and institutions, not abstract metrics.

You will also be working at a formative stage of the agri-tech ecosystem, where core platforms, standards, and practices are still being shaped. This is an opportunity to help define how technology supports agriculture at scale, by building foundational systems others will depend on.

This is a place for engineers who want to own outcomes, care deeply about correctness under uncertainty, and value long-term impact alongside sustainable velocity.