ML Engineer - On-Device Models

HafizPrime is a remote-first company developing Quran memorization technology designed to support Islamic learning. The platform combines a Flutter mobile application, a WordPress/PHP web interface, and Go-based backend infrastructure to deliver a comprehensive learning experience.

The company operates on a principle of long-term craftsmanship and ethical alignment with its Islamic mission. It explicitly does not pursue growth hacking, engagement manipulation, virality engineering, or data commercialization — focusing instead on building technology that genuinely serves learners.

The Role

As an ML Engineer (Senior level), you will own and evolve two core AI systems

• The recitation verifier: machine learning models that accurately check Quran recitation accuracy
• The retention score model: predictive models that identify what learners are likely to forget, enabling personalized review

Core Requirements

• Expertise in on-device machine learning (Core ML for iOS, NNAPI for Android, TensorFlow Lite)
• Demonstrated experience with speech audio processing and analysis
• Strong software engineering fundamentals

Preferred

• Background in or familiarity with Arabic phonetics

Work Arrangement

• Fully remote, any time zone worldwide
• 4-hour daily core overlap window for collaboration
• Full-time position

Compensation & Benefits

• Competitive salary benchmarked to your local market (60th–75th percentile)
• Equity available for senior and long-term contributors with transparent grant structure
• Unlimited PTO with a guaranteed minimum of 20 days per year
• Hajj leave: up to 6 weeks paid, once per tenure
• Ramadan flexibility: reduced core hours, no release pressure during Ramadan, extended Eid holidays
• Healthcare stipend tailored to your country
• Learning budget: $2,500 per year (can include classical Islamic studies courses)
• Home office stipend: $2,000 at hire + $500 annual refresh

Hiring Process

The company uses a transparent, fair hiring process

1. Written application with role-specific questions (substantive ~300-word responses expected)
2. Async video interview: record 30-minute responses to three provided questions (no scheduling required)
3. Paid work trial: real project work over 5–10 hours at your standard rate (replaces coding interviews)
4. Two structured conversations: one with the hiring manager, one with a peer engineer
5. Final decision within 14 days of trial completion

The company does not use LeetCode exercises, does not assign unpaid evaluations, does not use prolonged take-home tests (capped at 10 hours), and communicates a decision to all applicants — no ghosting.

How to Apply

Submit a portfolio and substantive application via the company's contact page, specifying this role in the subject line.