Machine Learning Engineer

We’re seeking a hands-on ML Engineer to transform cutting-edge research

into production features. You will own Tiny LLM on-device inference,

predictive analytics, dynamic escalation workflows, gamified modules, and

localisation pipelines—working end-to-end from model training to mobile

deployment.

Required Qualifications:

• Bachelor’s or Master’s in Computer Science, Engineering, or related field

• 2+ years’ hands-on experience with transformer architectures, fine-tuning,

and model inference

• Strong proficiency in Python, including libraries/frameworks such as

PyTorch, TensorFlow, sciki t- learn, and fastAPI

• Proven track record deploying ML models to production (TorchServe, ONNX,

HuggingFace Inference API)

• Solid data engineering skills: data lakes, batch pipelines, structured logging

(e.g., Airflow, Spark)

• Familiarity with edge/embedded ML: quantisation (4–8 bits), memory

footprints (20–50 MB), RAM budgets (100–300 MB)

• Experience configuring API Gateways, server-less functions, and message

queues (Kafka, Celery)

• Deep understanding of data security, privacy regulations (GDPR/HIPAA-

inspired), consent flows, and audit logging

• Expertise in localisation and NLP: neural translation, dialect adaptation,

multi-modal (text & voice) processing

• Comfortable in Agile/CI-CD environments with containerised micro-services

(Docker, Kubernetes)

Key Responsibilities:

• Develop and optimise Tiny LLM inference pipelines.

• Implement dynamic risk-based escalation workflows (sentiment 0.0–1.0;

thresholds 0.3–0.7; horizon 3–14 days)

• Build gamification engines (points: 10–100; streaks: 3–30 days; quest

windows: 1–7 days) to boost retention

• Integrate neural machine translation with regional dialect support (latency:

100–300 ms; BLEU: 30–50) for text and voice interfaces

• Architect offline data synchronisation (intervals: 1–24 hrs; payload: 5–50 kB)

and ensure seamless async sync under < 50 kB/s bandwidth

• Deploy models and services using TorchServe, ONNX, or HuggingFace

Inference API, and manage server-less scaling, API Gateway, Kafka/Celery

queues

• Collaborate with backend and mobile teams to meet performance targets (UI

load: 100–300 ms; battery drain: 1–3 %/hr)

• Embed security and compliance: AES-256 & TLS 1.3 encryption, consent

management, legal disclaimers, audit-grade logging

• Maintain high availability (99.9 % SLA), automated retraining cycles (1–4

weeks), and structured logging for analytics

• Write clean, production-grade Python code to support data pipelines, model

training, inference, and integration.

Preferred Skils:

• Prior work in digital health or mental-wellness applications

• Familiarity with mobile frameworks (React Native, Flutter)

• Experience designing or measuring gamification metrics

• Knowledge of federated learning or privacy-preserving ML

What We Offer:

- Competitive contract rate

- Remote work arrangement

- Opportunity to work on exciting projects with a talented team

If you’re passionate about building humane AI that transcends infrastructure

barriers and delivers personalised, proactive mental care, we’d love to hear

from you. Please share your resume and a brief note on a production ML system you’ve delivered end-to-end.