Filip's Solutions

Professional Experience

With over six years of experience in the complete software development life cycle. I’m developing well-tested code infrastructure in accordance with design patterns and best practices. I am capable of continuously delivering scalable solutions that ensure robust and optimized performance in cloud environments. My background in mathematics enhances my ability to innovate and implement practical AI-powered products particularly with Deep Learning.

Build a system which analyse videos from customers and give them advices how to improve their presentation based on interactive challenges.

Intelligent chatbot helping users make personalised retirement decisions using ML classification and conversational AI. Reduced need for manual financial consultations.

Multi-agent tool that autonomously generates Python code and documentation. Achieved top 5 score in the SWE-Bench benchmark, enabling developers to supervise rather than implement tasks.

Automated system converting handwritten container inspection reports into categorised digital data, enabling precise damage assessment and automated cost estimation.

Auto-detects and classifies brand logos from sports event videos. Achieved 93% mAP in logo detection and 91% accuracy in brand classification using metric learning and YOLO.

Intelligent assistant for real estate brokers automating lead interactions and suggesting next best actions to improve deal success using LLMs.

Agreement quality check application for a Fintech firm, automating contractual document analysis via a user-friendly interface using LLMs and RAG.

Evaluated various LLMs and RAG systems to establish a performance and quality benchmark. Compared retrieval strategies, model outputs and scoring metrics.

Built a BERT-based deep learning system to score event novelty, topic classification, and sentiment. Developed the full ML pipeline including training, deployment, performance monitoring, data modeling, and BI reporting for visualization.

Deep marketing analysis and develops ML models to assess which ads will be profitable.

Recommends specialists to pursue customers' personal goals like quitting smoking or losing weight.

Risk assessment and loan platform for the UK market. Delivered well-tested backend infrastructure in a close collaboration with the client's CTO.

Advanced trading service combining RL agents (DQN/PPO) with sentiment analysis and statistical strategies to forecast market prices and execute trades, achieving ~30% average annual return.

Finds the most suitable people for a given project, forecasting familiar technologies and calculating team compatibility index.

Calculates transport prices and forecasts costs and wear of parts based on web scraped data.

Certificates

I actively participate in certification courses to extend my knowledge in exciting fields of study.

Deep Learning Specialization
Convolutional Neural Networks
MySQL for Data Analytics and Business Intelligence
Big Data, Hadoop and Map Reduce
REST API Development
Docker Master Class for Machine Learning and Data Science
Neural Networks and Deep Learning
Hyperparameter Tuning, Regularization and Optymization
Structuring Machine Learning Projects
Convolutional Neural Networks
Sequence Models
Deep Learning Specialization
Convolutional Neural Networks
MySQL for Data Analytics and Business Intelligence
Big Data, Hadoop and Map Reduce
REST API Development
Docker Master Class for Machine Learning and Data Science
Neural Networks and Deep Learning
Hyperparameter Tuning, Regularization and Optymization
Structuring Machine Learning Projects
Convolutional Neural Networks
Sequence Models
Deep Learning Specialization

Strengths

Data Engineering

DBT
BigQuery
SQL
NoSQL
MySQL
PostgreSQL
MongoDB
ChromaDB
Qdrant
Elasticsearch
Airflow
PySpark
Pandas
Numpy
ETL
Parquet
Pub/Sub
Tableau
Power BI
Streamlit

Cloud

AWS
GCP
Azure
AWS EC2
GCP Compute Engine
AWS RDS
AWS Lambda
AWS EMR
AWS S3
GCP Storage
GCP BigQuery
GCP Firestore
AWS Sagemaker
AWS Bedrock
AWS Elastic BeanStalk
AWS Comprehend
AWS Rekognition
AWS SNS
AWS SQS
AWS CloudWatch
Azure MLStudio
Sentry
Datadog

Software Engineering

Django
Flask
FastAPI
Docker
Docker-Compose
Kubernetes
Swarm
NginX
CI/CD
Git
Linux
REST
OpenAI
LangChain
LangGraph
Redis
Celery
Multiprocessing
Asyncio
Big O
Architecture
DDD
UML
Poetry
Pre-commit
Prospector
Profiling
R
C++
YAML
Makefile
Pytest
UnitTest
TDD
E2E
Gherkin
Postman
Selenium
Stripe
DevOps
React

Machine Learning

NLP
LLM
RAG
Deep Learning
Reinforcement Learning
Computer Vision
PyTorch
Tensorflow
Keras
Hugging Face
Transformers
Scikit-learn
XGBoost
YOLO
Neural Networks
RNN
LSTM
BERT
Word2Vec
ARIMA
Anomaly Detection
PCA
OCR
Supervised
Unsupervised
Clustering
Regression
Random Forest
Ensemble
Association Rules
SVM
Recommender
Statistics
Mathematics
NLTK
MLFlow
DVC
Trulens
Ragas
SciPy
Matplotlib
Seaborn

Pet projects

A collection of personal and open-source projects where I experiment freely - exploring new tech stacks, sharpening algorithms, and turning ideas into working software.

Data Science

Machine Learning Engineer

MLOps Engineer

I feel satisfied using my Mathematical background in practical AI powered products.