I work on multi-agent LLM systems, evolutionary search for alpha signals on financial markets, generation of 2D and 3D apartment layouts, and medical image segmentation. I design the architecture and train the models these systems run on.
PhD candidate in Computational and Data Science and Engineering at an AI research institute. Research Engineer at Fusionbrain Lab. Co-founder of Fiber Pipe.
Evolutionary algorithms. LLM-guided evolution over program spaces, where a population of candidate programs is mutated, scored and selected across generations.
Multi-agent LLM systems. Multi-agent systems for applied tasks: role decomposition, orchestration and tool use, agent prompting, graph RAG, and training of the agent models themselves. Mid-training on domain corpora, post-training with SFT, LoRA and QLoRA, DPO and GRPO.
Generative models. Diffusion models and GANs for image synthesis and editing, including latent-space inversion and subject personalization.
Multimodal LLMs. Adapting vision-language models to new tasks with parameter-efficient fine-tuning (LoRA, QLoRA), supervised fine-tuning and reinforcement learning post-training (DPO, GRPO).
Evolutionary alpha search on markets. An LLM proposes and mutates factor expressions for Russian equities, generation after generation. Every candidate is scored by walk-forward backtests over a decade of daily data and screened against multiple testing with the deflated Sharpe ratio, probability of backtest overfitting and the model confidence set.
Multi-agent floor plan generation. Multi-agent LLM system that recommends furniture selections and placements within defined room boundaries, then renders realistic 3D rooms.
StyleGAN-2 encoder optimization. Encoder optimization with an analysis of image inversion and editing methods, raising the quality of generated and edited images.
Multimodal LLM fine-tuning. Parameter-efficient fine-tuning, including QLoRA, applied to InternVL2-1B to lift performance on the temporal tasks of MVBench.
SegMed. MedSAM segmentation of brain tumors on the LGG MRI dataset, measured against U-Net, U-Net++, PAN and DeepLabV3+ baselines and reaching mean IoU 0.644, then extended with FeatUp for feature upsampling and GAFL for adaptive frequency filtering.
Personalized image generation. Embeddings tuned through Textual Inversion in Stable Diffusion, creating custom tokens for personalized subjects.
Panorama stitching. SIFT and ORB feature matching with camera calibration, producing high-precision panoramic images.
Machine Learning-Driven Compensation for Non-Ideal Channels in AWG-Based FBG Interrogator I. A. Kazakov, I. V. Kulichenko, E. E. Kovalev, A. A. Treskova, D. D. Barma, K. M. Malakhov, I. V. Oseledets, A. V. Shipulin IEEE Sensors Letters, 2025
An exponential-regression calibration of a photonic integrated AWG interrogator reaches 3.17 pm RMSE against 7.11 pm for the segmented analytical model, and holds accuracy below 5 pm across an extended 2.9 nm span without refitting.
Generative image models for augmenting the training data of a face detector N. A. Andriyanov, Ia. V. Kulichenko Neurocomputers: Development, Application, No. 5, 2023. Listed by the Higher Attestation Commission. radiotec.ru Применение генеративных моделей изображений для аугментирования данных обучения детектора лиц.
A study of metric algorithms for face recognition Ia. V. Kulichenko, D. S. Utkin, A. C. Fan, N. I. Matuskov, I. M. Lopatkin, N. A. Andriyanov Popov Society for Radio Engineering, Electronics and Communications, 2023. eLIBRARY Исследование метрических алгоритмов в задаче распознавания лиц.
Faster face detection and identification through a motion detector D. S. Utkin, Ia. V. Kulichenko, N. A. Andriyanov Popov Society for Radio Engineering, Electronics and Communications, 2023. eLIBRARY Повышение скорости детекции и идентификации лиц на основе детектора движения.
Generative AI agents. Gitex Global 2025, Dubai World Trade Centre, 13 to 17 October 2025. gitex.com
Machine Learning-Driven Compensation for Non-Ideal Channels in AWG-Based FBG Interrogator. IEEE SENSORS 2025, Vancouver, Canada, 19 to 22 October 2025, presented online. ieee-sensorsconference.org
State of the Art YOLO Model in Object Recognition Tasks. XIV International Scientific Student Congress, Moscow, 13 to 24 March 2023. Second place in the competition.
Modern Text-to-Image Generation Technologies. International Scientific and Practical Conference of Students and Postgraduates, Moscow, 18 April 2023.
Three registered items with Rospatent, filed as Куличенко Яна Владимировна.
RU 2861310 C1, patent for invention, registered 4 May 2026. Method for interrogating fiber Bragg gratings through an arrayed waveguide grating demultiplexer on a photonic integrated circuit, with machine learning used to recover the reflected wavelength from per-channel optical power.
RU 2026613062, software registration, registered 3 February 2026. Software that converts the analog photodiode signal to digital, averages readings over a time window and computes measurement error including the standard deviation of the incoming voltage level. Written in C++.
RU 2025681052, software registration, registered 11 August 2025. Interrogator user software for real-time acquisition, processing and display of fiber Bragg grating sensor data, with adaptive analysis and storage of historical values. Written in TypeScript.
Methods: agent orchestration and tool use, graph RAG, mid-training, SFT, LoRA and QLoRA, DPO, GRPO, genetic programming, MAP-Elites, walk-forward validation.
Co-founder of Fiber Pipe, which applies computer vision and fiber-optic sensing to pipeline monitoring in the Arctic. First place at the EnergyTechnoHub incubator in Saint Petersburg, winner of the Triple Point pitch competitions, supported by research grants.
PhD in Computational and Data Science and Engineering, 2025 to 2028, AI research institute. MSc in Data Science, 2023 to 2025. BSc in Applied Mathematics and Informatics, 2019 to 2023.
Open to research collaboration on generative AI, multi-agent systems, evolutionary algorithms and alpha signal search on financial markets.
