Feliks Bańka

Feliks Bańka

AI Engineer & Research Engineer

About Me

I am a master’s student in Artificial Intelligence at Warsaw University of Technology and an AI Engineer at AI Clearing, where I build production LLM systems, RAG pipelines, and multi-agent architectures for complex analytical workflows. My research focuses on multi-agent systems, LLM evaluation and verification, game-theoretic decision systems, and reliable AI-generated outputs. My recent work studies bounded verification, benchmark-grounded evaluation, and explanation-aware coordination, alongside research in cooperative MARL and quantitative AI. I have authored and co-authored work accepted at venues including ICCS, PACIS, ACIIDS, ICAART, and AAMAS workshops, with current manuscripts under review or submitted to AAAI 2027, NeurIPS 2026, ICTAI 2026, and ICADL 2026.

Education

Master's degree • Warsaw University of Technology

Sep 2025 – Feb 2027

Major: Artificial Intelligence • GPA: 4.60/5.0

Thesis: Investigating the Application of Game Theory in LLM-based Multi-Agent Systems

B.Eng. • Warsaw University of Technology

Sep 2021 – Jun 2025

Major: Automatic Control & Robotics

Thesis: AI Architectures for Option Pricing and Portfolio Management: Transformers to MAS

Selected Publications & Manuscripts

Passing an Audit Is Not Enough: Claim Composition and Informed Inspection for LLM Outputs

AAAI 2027 (Under Review)

Studies bounded audits of LLM outputs, including claim composition, generator-reviewer incentives, and informed inspection of decision-critical claims.

ClaimReplay: Harness for Evaluating LLM-Generated Documents

ICTAI 2026 (Submitted)

Introduces a benchmark-grounded evaluation harness for replayable analysis of LLM-generated documents and selective verification policies.

More Claims, Less Evidence: Bounded Verification of AI-Generated Digital Knowledge Artifacts

ICADL 2026 (Submitted)

Examines how verification capacity and reliability change when AI-generated artifacts contain more claims than a bounded reviewer can inspect.

Breaking Advantage Symmetry: Structured Baselines for Cooperative MARL

NeurIPS 2026 (Under Review)

Proposes a PPO-compatible critic design with a shared team baseline and zero-mean agent-specific residuals to improve credit assignment in cooperative multi-agent reinforcement learning.

MultiHedge: Adaptive Coordination via Retrieval-Augmented Control

ICCS 2026

Introduces a retrieval-augmented multi-agent framework for portfolio decision support, combining historical precedent retrieval, structured allocation decisions, and risk-aware evaluation.

Toward Explanatory Equilibrium: Verifiable Reasoning as a Coordination Mechanism under Asymmetric Information

AAMAS 2026 (Workshop)

Studies explanation-aware multi-agent coordination under asymmetric information and models LLM communication as a signaling problem with verifiable reasoning.

DeltaHedge: A Multi-Agent Framework for Portfolio Optimization

PACIS 2025

Presents a multi-agent reinforcement learning framework for portfolio optimization and options-based hedging under changing market conditions.

Options Pricing Platform with NNs, LLMs & RL

ACIIDS 2025 • Best Student Paper Award

Integrated neural networks, volatility modeling, sentiment analysis via LLMs, and reinforcement learning into a hybrid option-pricing and trading platform.

Transformer Architectures for Option Pricing and Forecasting

LNAI (Extended ICAART)

Compares Transformer-based forecasting architectures with classical and deep sequence baselines across equities, indices, and crypto options.

Applying Informer for Option Pricing

ICAART 2025

Adapts the Informer architecture to option pricing and studies long-sequence modeling for derivatives forecasting.

Experience

AI Engineer • AI Clearing

Jul 2025 – Present

Architected and deployed a production multi-agent LLM system using LangGraph and LangChain, including stateful workflows, tool routing, backend API calls, and RAG. Built async FastAPI services with strict Pydantic v2 schemas for reliable structured outputs and downstream integrations, and designed retrieval pipelines over heterogeneous enterprise data using OCR, vision LLMs, BM25, reranking, vector search, and selective context injection.

Impact: Increased benchmark accuracy from 78% to 98% through evaluation, failure analysis, and regression testing. Reduced latency from 80s to 20s and token usage from 1.1M to 200k through caching, selective routing, and context optimization.

Developer • Regionalne Przysmaki

Oct 2023 – Jun 2025

Developed Python-based automation tools and data pipelines. Analyzed market trends and optimized strategies to support business operations.

Full Stack Developer • Freeport Metrics

Jul 2023 – Oct 2023

Built React (TypeScript) and .NET micro-frontends, implemented REST APIs, and integrated backend logic in cross-functional Agile teams.

Selected Projects

Avellaneda–Stoikov Market Making Simulator

Python • Quantitative Research Engineering

Built a research-oriented market-making simulation framework based on inventory-aware quoting, partial fills, fees/slippage, risk controls, calibration utilities, and reproducible strategy benchmarking.

Short-Horizon Order-Flow Alpha Benchmark

Python • Market Microstructure

Developed an event-level order-book research pipeline for short-horizon mid-price prediction under latency, transaction cost, and anti-leakage constraints, with causal features, chronological validation, and post-cost trading diagnostics.

Extracurricular Activities

Co-founder & President • KN Intelligent Information Systems Nexus

Jun 2025 – Present

Leading a student research group focused on generative AI, multi-agent systems, LLM applications, and argument modeling. Promoting practical AI applications in finance and decision systems.

Vice President • KN Business in Technology

Feb 2023 – Jun 2024

Led AI/LLM-oriented student projects, organized tech-business meetups, and managed a 10-member team delivering workshops and hackathons in Warsaw.

Member • Enactus WUT

Nov 2021 – Sep 2024

Collaborated on entrepreneurship projects addressing social challenges through innovation, with emphasis on leadership and cross-functional teamwork.

Honors & Awards

Best Student Paper Award (ACIIDS 2025)

Top-ranked student paper by the program committee.

EU / Ministry of Science Research Grant

FERS 2021-2027

Individual research funding.

Minister of Science Scholarship

National award granted to 400 top students nationwide.

Skills & Tools

ML / AI

PyTorch Transformers Reinforcement Learning LLMs RAG Multi-Agent Systems Game Theory

LLM Systems & Evaluation

LangChain LangGraph Hugging Face Structured Outputs Benchmarking Evaluation Failure Analysis

Data & Retrieval

Qdrant Embeddings BM25 Reranking Neo4j Knowledge Graphs OCR Vision LLMs

Engineering

Python TypeScript FastAPI Async / Concurrency Task Queues Docker Kubernetes Caching

Research

Research Design Literature Review Experimental Design Ablation Studies Statistical Analysis Academic Writing Peer Review Reproducible Research

Quantitative Finance

Options Pricing Portfolio Optimization Market Making Order-Flow Alpha Backtesting Risk Metrics Market Microstructure

Languages

Polish • Native English • Advanced German • Intermediate

Hobbies & Interests

⚽️
Football Enthusiast
🎵
Songwriting & Music
📈
Stock Market Analysis
🤖
DIY Robotics Projects