Selected engineering work

Projects built to solve real engineering problems.

I keep this portfolio deliberately focused. These six projects represent the work I can explain deeply: the problem, architecture, implementation decisions, failure modes, and what I learned while building them.

The strongest themes across the work are agentic AI, multi-agent systems, backend and distributed architecture, asynchronous processing, observability, and grounded AI workflows.

Flagship work

Research depth, agent architecture, and applied AI engineering.

Flagship Research · Distributed AI

STRATON-LLM

An ontology-based communication and semantic negotiation system for heterogeneous intelligent agents that do not share the same terminology or prior knowledge of one another.

Designed and implemented end to end, including distributed agent communication, service discovery, negotiation workflows, persistence, and LLM-assisted semantic reasoning.

PythonFastAPIRabbitMQConsulLLMsMulti-Agent Systems
Completed system · further validation in progressView case study →

Open Multi-Agent Framework

LATTICE Framework

An open, domain-agnostic multi-agent framework for decentralized collaboration between specialized autonomous agents.

v0.1 is design-complete and implementation is pending. The architecture defines parent-driven routing, broker-mediated communication, explicit state ownership, memory and knowledge boundaries, failure handling, and first-class observability.

PythonAsyncIOMulti-Agent SystemsMessage BrokerLLM Agents
Design complete · implementation pendingView case study →

Applied AI · Personalization Platform

DeepFeed AI

A personalized knowledge-discovery platform combining content discovery, explainable ranking, feedback-driven preference learning, and agentic research adaptation.

The repository contains a FastAPI backend, Next.js frontend, PostgreSQL/pgvector persistence, RabbitMQ/Celery processing, explicit recommendation traces, adaptive agents, observability, automated tests, and CI.

PythonFastAPIPostgreSQLpgvectorRabbitMQCeleryOpenTelemetry
Active developmentView case study →

Supporting engineering work

Grounding, orchestration, and real-time system design.

Grounded AI · Research Engineering

Research-Paper Deconstruction Agent

A two-stage system that reconstructs research PDFs into structured Markdown/JSON and then runs a LangGraph multi-agent workflow for grounded technical analysis.

Implements layout-aware preprocessing, semantic paper-structure classification, five research agents, separate faithfulness verification, deterministic hallucination scoring, unsupported-claim removal, and experiment-reproduction analysis.

PythonLangGraphPyMuPDFGrounded AIPDF ProcessingFaithfulness

Working · 5 of 8 analysis phases implemented

View case study →

Conversational AI · Stateful Backend

Agentic Chatbot

A multi-user conversational AI backend that turns stateless LLM calls into persistent, memory-aware conversations with streaming, conflict resolution, and evaluation.

Implements a separate auth gateway, JWT-scoped conversations, three-layer memory, pending memory conflicts, SSE streaming, LLM self-evaluation, PostgreSQL persistence, rate limiting, OpenTelemetry tracing, and Docker Compose deployment.

PythonFastAPIPostgreSQLSSEOpenTelemetryJaegerJWT

Working system

View case study →

Software Engineering · System Design

Real-Time Collaborative Document Platform

A collaborative document system designed around two or more users editing shared text in real time while preserving document state across sessions.

Explores real-time communication, operation propagation, persistence, service boundaries, and the architectural trade-offs behind collaborative editing systems.

WebSocketsPostgreSQLReal-Time SystemsDistributed SystemsSystem Design

Architecture / portfolio implementation

How I present projects

The goal of each case study is not to list tools. It is to document the problem, constraints, architecture, key decisions, implementation, testing or evaluation, and the trade-offs that shaped the system. STRATON-LLM currently has the deepest case study; the remaining project pages will be expanded progressively.

Want to discuss the engineering?

I'm open to Software Engineering, AI Engineering, Applied AI, Agentic AI, and AI/backend opportunities.

Get in touch →