MEng Computer Science at Durham University · AI / Systems / Agentic Networks

I'm a MEng Computer Science student at Durham University building agentic AI systems and intelligent tooling. My work focuses on bridging machine learning, systems engineering, and real-world deployment, from multi-agent optimisation to computer vision pipelines. I'm currently interested in part-time roles and joining passionate teams working on interesting projects and startups.

AI AgentsSystems EngineeringComputer Vision

Capabilities

Agentic AI Systems

Build multi-agent workflows, tool-using agents, retrieval pipelines, and AI automations that connect models to real tasks.

LangGraphAutoGenMCPRAGChromaDBOllamaAWS BedrockWatsonX

Backend & Systems

Build APIs, services, pipelines, and performance-conscious systems with an emphasis on reliability, structure, and real deployment constraints.

PythonNode.jsC++REST APIsAWS LambdaDistributed SystemsSystems DesignPerformance Engineering

Applied ML

Comfortable taking ML work from analysis to implementation, including training, evaluation, forecasting, and multimodal pipelines.

scikit-learnNumPyComputer VisionNLPTime SeriesRandom ForestDimensionality ReductionData Analysis

Delivery & Collaboration

Used to working in teams, shipping project work, writing documentation, and communicating clearly with both technical and non-technical stakeholders.

GitJestDocumentationSprint PlanningClient WorkLeadershipTestingTeam Collaboration

Projects

  • Built dynamic web pages using DOM manipulation and AJAX to load JSON data.
  • Designed and deployed a REST API server in Node.js to handle JavaScript requests efficiently.
  • Improved cross-browser compatibility with ESLint integration and comprehensive Jest test coverage.
  • Collaborated in a team of 7 to deliver a full-stack AI-powered educational game to an IBM client.
  • Built an agentic reasoning system using LangGraph and WatsonX embeddings with a locally hosted IBM Granite model.
  • Integrated ChromaDB vector storage to inject game rules and quiz content into agent reasoning.
  • Produced industry-style documentation including requirements, sprint plans, and testing reports.
  • Implemented a distributed MCP runtime for real-time KPI optimization in a 6G-aligned telecom simulator.
  • Engineered Strands SDK agents combining reinforcement learning and tool-based reasoning for adaptive control.
  • Built a scalable Lambda-MCP pipeline enabling parallel agent execution and about 20% simulated power savings.
  • Designed a multi-agent pipeline using Microsoft AutoGen for financial and company intelligence.
  • Built a web scraper to extract, filter, and semantically analyze financial news.
  • Integrated metrics into an LSTM model to predict company performance trends.
  • Performed statistical analysis to identify skew, outliers, and class structure in multimodal datasets.
  • Implemented custom Random Forest classifiers using entropy and Gini impurity splits.
  • Applied dimensionality reduction to improve generalization, achieving about 93% accuracy on unseen data.
  • Implemented custom hash tables and probabilistic data structures for fast membership and lookup queries.
  • Focused on cache locality, memory layout, and branch prediction to maximize throughput.
  • Benchmarked performance against standard library implementations to validate improvements.

Experience

Activity Manager

Oxford International

Jun 2025 - Aug 2025

  • Led a team of Activity Leaders delivering onsite programmes to international students.
  • Managed staff rotas and responsibilities to ensure smooth daily operations.
  • Oversaw equipment, resources, and communications for large group activities.
LeadershipOperationsManagement

Activity Leader

Oxford International

Jun 2023 - Aug 2023, Jun 2024 - Aug 2024

  • Led excursions and supervised groups of up to 70 students across the UK.
  • Handled high-pressure and safety-critical situations with responsibility and adaptability.
  • Worked collaboratively with staff to deliver engaging, well-organized activities.
LeadershipTeamworkResponsibility

MEng Computer Science Student

Durham University

Oct 2023 - Jul 2027

  • On track for First Class Honours (1:1).
  • Focus areas include AI engineering, distributed systems, and machine learning systems.
AISystemsML

Contact.

MEng Computer Science at Durham University · AI / Systems / Agentic Networks

Availability

Open to internships, part-time roles, and project-based work in AI engineering, backend systems, and agentic tooling. During university term time I can usually commit 5-10 hours per week remotely. From June 2026 onward I'll have much broader availability for summer internships and higher-hour commitments. Based in the UK and open to remote or hybrid roles. I'm also happy to relocate internationally for the summer for the right opportunity.

balatssasha@gmail.com->
GitHubLinkedIn
LocationUnited Kingdom
AvailabilityOpen to opportunities