Senior ML Engineer

We are looking for Senior ML Engineers with deep experience in computer vision / Anti Fraud/ biometric/ identity verification systems and building production-grade solutions.

Compensation: the offer depends on the expertise

We are looking for an engineer who:

• worked with face Anti Fraud / KYC / identity verification

• understands the architecture of enterprise biometric systems

• can build production-level ML/CV pipelines

• has experience in anti-spoofing, liveness detection, face verification

Key advantages (strong advantage)

• Experience in biometrics KYC company

• Participation in NIST FRVT evaluations

• experience in preparing solutions for iBeta Level 2

• passing or participating in the iBeta Level 2 certification

• Understanding biometric compliance & certification standards

• Experience in deepfake detection / advanced anti-spoofing

Technical requirements

• Strong ML / Deep Learning background

• Computer Vision production experience

• Python + PyTorch / TensorFlow

• training / fine-tuning CV models

• end-to-end ML systems design

• deployment & optimization in production

• working with large-scale datasets

• latency / inference optimization

• GPU optimization experience

Mandatory requirement (AI-first engineer)

Active use of AI tools in work:

• ChatGPT / Claude / Gemini

• Cursor / Windsurf / GitHub Copilot

• AI-assisted coding & debugging

• prompt engineering

• AI-driven research workflows

• synthetic data generation

• LLM API integrations

• TensorRT / CUDA / ONNX Runtime

• OpenCV / InsightFace / DeepFace

• mobile / edge optimization

• OCR / document verification

• AWS / GCP

• Docker / Kubernetes

• MLOps / CI/CD

We especially appreciate

• implementation of SOTA papers in production

• understanding the difference between demo vs production AI

• independent technical decisions

• the speed of adaptation in a fast-moving environment

Development and improvement of Core CV/Biometric systems

• Anti-fraud and spoofing: Develop, train and implement advanced models for liveness detection, anti-spoofing and deepfake detection to protect the system from fakes (photos, videos, masks).
• Face Recognition and verification: To design and optimize face verification and identification algorithms for KYC (Know Your Customer) and Identity Verification systems.
• Document Recognition: Develop and integrate solutions for OCR (text recognition) and document authentication.

Architecture design and Production-grade pipelines

• End-to-End development: Design the architecture of Enterprise Biometric Systems from data collection to deployment.
• Building CV pipelines: Create fault-tolerant, scalable, and fast ML/CV pipelines that are ready for high loads (rather than just running as demo scripts).
• Working with data: Collect, mark up, and prepare large-scale datasets, as well as generate synthetic data for model training.

Optimization and MLOps (Speed and Efficiency)

• Inference & GPU optimization: Reduce latency and optimize the resource consumption of models using TensorRT, CUDA, ONNX Runtime.
• Edge & Mobile: Adapt and optimize heavy models to work on mobile and edge devices.
• Deployment and infrastructure: Containerize solutions (Docker/Kubernetes), deploy them in the clouds (AWS/GCP), and build MLOps processes (CI/CD for ML).

Certification and Compliance with Standards (Compliance)

• Preparation for iBeta Level 2: Prepare the company's biometric solutions and architecture for successful completion of the international iBeta Level 2 certification.
• Test Participation: Prepare models and participate in independent global biometrics assessments such as NIST FRVT evaluations.
• Maintain biometric systems in accordance with international compliance and security standards.

AI-First approach and Research

• Acceleration through AI: Build your own development workflow, actively using AI assistants (Cursor, Windsor, Copilot, ChatGPT/Claude), prompt engineering and LLM API for fast code writing, debugging and reserching.
• Transferring SOTA to Production: Monitor the latest scientific articles (SOTA papers) on Computer Vision / Biometrics and promptly implement the best approaches into a real product, understanding the difference between a beautiful academic demo and a stable enterprise solution.

We offer:

- Work for a biometric IT company

- Comfortable office

- Open communication in the team

- Offline work from the office on schedule 5/2 from 10:00 to 19:00, at the end of the probation period (3 months) there is an opportunity to discuss a hybrid work format