Shahab Anbarjafari - Lead AI Scientist — Computer Vision & Sensor Fusion

Q: Where is Shahab based?

Shahab is based in Veibri, Estonia .

Q: What languages does Shahab speak?

Shahab speaks the following languages: Persian (Native), English (Advanced), Turkish (Advanced), Estonian (Elementary) .

Q: How many years of experience does Shahab have?

Shahab has at least 13 years of experience. During this time, Shahab has worked in at least 11 different roles and for 14 different companies . The average length of individual experience is 1 year and 11 months . Note that Shahab may not have shared all experience and actually has more experience.

Q: What is Shahab's latest experience?

Shahab's most recent position is Lead AI Scientist — Computer Vision & Sensor Fusion at Boston HealthTech .

Q: What companies has Shahab worked for in recent years?

In recent years, Shahab has worked for Boston HealthTech , LanceSoft , Estonian Business School , Quilt LLC , and SimplifyWorkforce (USTech) .

Q: Which industries is Shahab most experienced in?

Shahab is most experienced in industries like Information Technology , Retail , and Automotive . Shahab also has some experience in Banking and Finance , Professional Services , and Education .

Q: Which business areas is Shahab most experienced in?

Shahab is most experienced in business areas like Research and Development , Product Development , and Business Intelligence . Shahab also has some experience in Information Technology , Project Management , and Finance .

Q: Which industries has Shahab worked in recently?

Shahab has recently worked in industries like Information Technology , Automotive , and Retail .

Q: Which business areas has Shahab worked in recently?

Shahab has recently worked in business areas like Research and Development , Product Development , and Information Technology .

Recommended expert

Veibri, Estonia

Experience

May 2025 - Sep 2025

5 months

Lead AI Scientist — Computer Vision & Sensor Fusion

Boston HealthTech

Designed multi-modal DL pipeline (vision, wearables, ambient sensors); synchronization, denoising, late-fusion classifiers.
Shipped the first MVP, deployed and tested with a healthcare center covering 100 patients; HIPAA-aware data handling.
Real-time activity recognition (falls, gait, ADL) on edge gateways; PyTorch temporal CNN/Transformer backbones.
Clinical evaluation harness, dashboards, alerting; FHIR adapters for EHR integration.

Nov 2024 - Aug 2025

10 months

Senior Director of AI/GenAI

LanceSoft

Built LLM-driven pipelines (CNN/RNN/GNN/GAT) supporting discovery workflows.
Stakeholder leadership, estimation, and team mentoring across bids.

Sep 2024 - Present

1 year 7 months

Visiting Professor

Estonian Business School

Research advising and curriculum development on GenAI & responsible AI.
Tooling: Python, R, SQL, MATLAB, PowerBI for academic/industry projects.

Aug 2024 - Apr 2025

9 months

ML Scientist

Quilt LLC

SLM fingerprinting, real-time summarization, phishing detection.
Python, NLTK/SpaCy, Llama/Gemini; SQL, MATLAB; TensorFlow; JS/HTML; Spark; AWS.

May 2024 - Jun 2025

1 year 2 months

Senior Director of AI/GenAI

SimplifyWorkforce (USTech)

Built CV/JD parsing, cover-letter generator, AI interviewer, video profiler; Responsible AI guardrails.
Stack: Python, TensorFlow, LangChain, RAG/GraphRAG, OpenCV, SQL, AWS, Spark, PowerBI; Twilio integrations.

May 2024 - Nov 2024

7 months

Lead ML Scientist — Graph Attention Networks & Process Optimization

DACH Pharma

Delivered 20% lower MAE on molecular property prediction, +18% AUC on drug–target interaction, and +15% synthesis yield prediction accuracy vs. baselines — helping to reach 25% faster lead identification.
Architected multi-task PyTorch Geometric GAT with multi-head attention, late graph readouts, and task-specific heads (regression/classification) over 50k molecules and 10k protein–ligand pairs.
Built RDKit-driven molecular graph featurization (atoms/bonds + global descriptors) and bipartite reaction graphs to model yields/byproducts; integrated Monte Carlo condition search for route optimization.
Shipped a production inference API that plugged into the client’s process simulator, enabling high-throughput screening (10k compounds/day) with seconds-level latency instead of hours.
Drove interpretability for regulatory/compliance via attention map visualizations highlighting sub-structures driving predictions; accelerated chemist review/triage loops.
Orchestrated pre-training on ChEMBL/PubChem then domain fine-tuning; established rigorous splits, early stopping, and PU-learning for DTI to curb false positives.
Operationalized on NVIDIA A100 cluster (50 epochs, 256-mol batches) with AdamW + dropout; delivered hardened pipelines, config-driven experiments, and tracking.
Partnered cross-functionally (computational chemistry & process engineering) to translate model insights into higher-yield routes and reduced byproducts at scale.

Mar 2024 - Jun 2025

1 year 4 months

AI SME

FlowStep

Generative image/text models (Stable Diffusion, GANs) and LLM APIs (GPT/Llama/Gemini).
Tooling: Python (Pandas/NumPy/Sklearn), PyTorch, OpenCV; Figma collaboration.

May 2023 - Aug 2024

1 year 4 months

AI Lead

LHV Pank

AML with Graph Attention Networks, compliance copilot, dynamic credit scoring.
Tooling: PyTorch/NetworkX, HF Transformers, ChatGPT API, SQL/Spark, PowerBI, AWS.

Sep 2022 - Aug 2023

1 year

ML Scientist

Black Wall (formerly BotGuard)

Click/mouse-movement behavior analysis; streaming ML at scale.
Python/Sklearn, TensorFlow; JS/HTML; SQL/Spark; AWS.

Sep 2020 - May 2024

3 years 9 months

Visiting Professor

Yildiz Technical University

Taught AI courses.
Consulted on smart city initiatives.
Tooling: Python/MATLAB/R/SQL; PowerBI for urban analytics; civic-tech collaborations.

Mar 2020 - Jul 2025

5 years 5 months

AI Scientist — GenAI & Responsible AI

PwC Finland

Orchestrated firm-wide GenAI programs in Finland and the Nordics: operating model, governance, model risk controls, and privacy-by-design; acted as a key driver of Responsible AI at the EMEA & Global level.
Built and productionized LLM & deep learning solutions: document/file parser, document content generator, audio-based chatbot, and document-matching algorithms; integrated with client systems and compliance workflows.
For a Nordic bank: founded the AI team and led 5 ML engineers & 7 data scientists; delivered AML, Transaction Monitoring, and KYC models/pipelines; initiated a group-wide ChatGPT Enterprise rollout and ran enablement for all lines of service.
Advised and delivered across Retail, DeepTech, FinTech, and Manufacturing: anomaly detection, predictive maintenance, computer vision for quality inspection, ESG & financial reporting automation, and a virtual assistant for internal users.
Led Nordic client AI engagements end-to-end: discovery → MVP → production; set standards for experiment tracking, CI/CD, observability, and risk assessments.
Stood up PwC FI Metaverse practice elements to explore spatial computing use cases aligned with enterprise AI roadmaps.
Stack: Python, PyTorch/TensorFlow, HuggingFace Transformers, SQL/Spark, Docker/Kubernetes, MLflow/Airflow, Power BI; deployments on AWS/GCP; security, auditability, and MLOps best practices baked in.

Nov 2019 - Sep 2024

4 years 11 months

Co-founder & CTO

Alpha3D

Developed 3D content creation using diffusion-based models for AR.
Tooling: Python with TensorFlow/PyTorch; GANs/Transformers/U-Net; OpenCV; JS/HTML; SQL; AWS/GCP.

Sep 2018 - Aug 2020

2 years

Discovery Search Team Lead

Rakuten Inc via iCV Lab (University of Tartu)

Led multimodal discovery search research for e-commerce, improving recommendations.
Python/PyTorch; CNN/RNN/Transformers; SQL/Spark; AWS; experimentation and metrics dashboards.

Sep 2013 - Apr 2024

10 years 8 months

Professor / Head of Intelligent Computer Vision (iCV) Lab

University of Tartu

Secured over 10 million euros in funding across 47 ML/CV/AI contracts, partnering with Mifundo, Jobbatical, Veriff, Clevon, AuveTech, OthersEyes, and Miros.AI (formerly Find.Fashion).
Veriff: developed core facial verification AI now supporting secure identity verification, including public-sector use in Estonia.
Clevon: built autonomous-traffic perception, including a traffic light status recognition system exceeding 85% accuracy.
AuveTech: designed LIDAR and RGB based tiny-object detection to improve autonomous mini-van safety and obstacle avoidance.
Jobbatical: created NLP pipelines for candidate and job data harvesting, enrichment, and domain-specific skill matching.
OthersEyes: pioneered first-impression analysis from images and video to help mitigate cyberbullying and promote online empathy.
Miros.AI: engineered the core AI for a recommendation engine enabling wordless search in e-commerce.
Delivered industry projects in facial verification and autonomous vehicle detection; led end-to-end research, prototyping, and transfer to production with partners such as Veriff and Clevon.
Tooling and platforms: NumPy, OpenCV, LIDAR/RGB computer vision, YOLO/SSD, SQL, MATLAB, R, Spark, Power BI, AWS, and GCP.

Industry Experience

See where this freelancer has spent most of their professional time.

Experienced in Information Technology, Retail, Automotive, Banking and Finance, Professional Services, and Education.

Information Technology

Retail

Automotive

Banking and Finance

Professional Services

Education

Business Area Experience

See which departments and functions this freelancer has contributed to most.

Experienced in Research and Development, Product Development, Information Technology, Business Intelligence, Project Management, and Finance.

Research and Development

Product Development

Information Technology

Business Intelligence

Project Management

Finance

Summary

Visionary AI leader and scientist with end-to-end ownership of agentic AI, Generative AI, Computer Vision, and NLP solutions—from research and prototyping to production deployments. Proven ability to build and lead high-performance teams, establish AI research labs, and deliver outcomes across startups, global enterprises, and consulting engagements. Deep hands-on experience with Azure/OpenAI, AWS, Hugging Face, Google Gemini, Claude; training & fine-tuning transformer and vision models; and MLOps on AKS/EKS/Vertex. Operates comfortably at exec strategy and implementation detail.

Skills

Ai/ml: Agentic Ai, Genai & Rag, Nlp, Computer Vision, Time Series, Reinforcement Learning (Practical), Evaluation/guardrails
Models/frameworks: Transformers, Gpt/llama/gemini/claude, Diffusion, Yolo/detectron2/mmdetection, Pytorch, Tensorflow/keras
Agents/orchestration: Langchain, Crewai, Autogen, Custom Tool-use & Evaluators
Cloud/mlops: Azure Ml/aks/adf, Aws Sagemaker/eks/lambda, Vertex, Docker/k8s, Mlflow, W&b, Ray
Data/infra: Lakehouse, Delta/parquet, Feature Stores, Faiss/milvus/pgvector, Kafka, Elastic
Engineering: Python, C++, Sql, Typescript/node, Fastapi, React/next.js

Languages

Persian

Native

English

Advanced

Turkish

Advanced

Estonian

Elementary

Education

Oct 2008 - Jun 2011

Eastern Mediterranean University

Ph.D., Electrical & Electronic Engineering, Computer Science Major · Electrical & Electronic Engineering

Oct 2007 - Jun 2008

Eastern Mediterranean University

M.Sc., Electrical & Electronic Engineering, Computer Science Major · Electrical & Electronic Engineering · 4.00/4.00

Oct 2003 - Jun 2007

Eastern Mediterranean University

B.Sc., Electrical & Electronic Engineering, Robotics Major · Electrical & Electronic Engineering · 3.97/4.00

Certifications & licenses

IEEE Senior Member

IEEE

Statistics

Experience

Total positions 14

Experience in Information Technology 12 y

Avg length 2 y 6 m

Longest experience 10 y 7 m

Expertise

Recent roles Lead AI Scientist — Computer Vision & Sensor Fusion, Senior Director of AI/GenAI, Visiting Professor

Main industries Information Technology, Retail, Automotive

Main business areas Research and Development, Product Development, Information Technology

Qualifications

Highest degree Doctorate

Certifications earned 1

Profile

Created

October 2025

Need a freelancer? Find your match in seconds.

Try FRATCH GPT

Frequently asked questions

Do you have questions? Here you can find further information.

Where is Shahab based?

Shahab is based in Veibri, Estonia.

What languages does Shahab speak?

Shahab speaks the following languages: Persian (Native), English (Advanced), Turkish (Advanced), Estonian (Elementary).

How many years of experience does Shahab have?

Shahab has at least 13 years of experience. During this time, Shahab has worked in at least 11 different roles and for 14 different companies. The average length of individual experience is 1 year and 11 months. Note that Shahab may not have shared all experience and actually has more experience.

What roles would Shahab be best suited for?

Based on recent experience, Shahab would be well-suited for roles such as: Lead AI Scientist — Computer Vision & Sensor Fusion, Senior Director of AI/GenAI, Visiting Professor.

What is Shahab's latest experience?

Shahab's most recent position is Lead AI Scientist — Computer Vision & Sensor Fusion at Boston HealthTech.

What companies has Shahab worked for in recent years?

In recent years, Shahab has worked for Boston HealthTech, LanceSoft, Estonian Business School, Quilt LLC, and SimplifyWorkforce (USTech).

Which industries is Shahab most experienced in?

Shahab is most experienced in industries like Information Technology, Retail, and Automotive. Shahab also has some experience in Banking and Finance, Professional Services, and Education.

Which business areas is Shahab most experienced in?

Shahab is most experienced in business areas like Research and Development, Product Development, and Business Intelligence. Shahab also has some experience in Information Technology, Project Management, and Finance.

Which industries has Shahab worked in recently?

Shahab has recently worked in industries like Information Technology, Automotive, and Retail.

Which business areas has Shahab worked in recently?

Shahab has recently worked in business areas like Research and Development, Product Development, and Information Technology.

What is Shahab's education?

Shahab holds a Doctorate in Electrical & Electronic Engineering from Eastern Mediterranean University, a Master in Electrical & Electronic Engineering from Eastern Mediterranean University and a Bachelor in Electrical & Electronic Engineering from Eastern Mediterranean University.

Does Shahab have any certificates?

Shahab has 1 certificate: IEEE Senior Member.

What is the availability of Shahab?

Shahab is immediately available full-time for suitable projects.

What is the rate of Shahab?

Shahab's rate depends on the specific project requirements. Please use the Meet button on the profile to schedule a meeting and discuss the details.