Selected web products and systems

Self-hosted OCR + Vision-Language Model pipeline running on GPU with CUDA optimization. Handles scanned documents, complex layouts, tables, and embedded images. Benchmarked multiple OCR engines (PaddleOCR, Nanonets, MistralOCR, Deepseek OCR, OlmOCR...) against a representative corpus to select the optimal engine per document type. Includes intelligent chunking with semantic embedding for downstream RAG.

Implementation: PaddleOCR, MistralOCR, Nanonets, Deepseek OCR, OlmOCR, Gemini Vision, CUDA, Python, Pinecone

Multi-agent RAG pipeline for complex legal document corpora. Custom AI chunker segments non-standardized legal text while preserving cross-references. Hybrid retrieval combines BM25 keyword search with vector similarity search, fused via Reciprocal Rank Fusion (RRF) for optimal ranking. LLM-as-Judge validation loop scores retrieved passages for relevance and completeness before answer generation. Metadata enrichment adds jurisdiction, document type, and date filters.

Implementation: LangGraph, Anthropic Claude, Pinecone, BM25, Vector Search, RRF, LLM-as-Judge, Metadata Enrichment, Python

Retrieval-based assistant for Health AI. Answers questions about health coverage, procedures, and documentation requirements by retrieving from official health documents - PPTX presentations, SQL databases, and regulatory text. Self-hosted infrastructure: Supabase for data and authentication, Qdrant for vector storage, OpenRouter for multi-model LLM access. Hugging Face embeddings with reranking pipeline. Arize Phoenix observability for tracing retrieval quality and latency. Chunked document retrieval with relevance scoring, source attribution on every answer, and memory for multi-turn conversations about complex healthcare procedures.

Implementation: Qdrant, Supabase, OpenRouter, Hugging Face, Reranking, Arize Phoenix, Python, RAG

Multilingual retrieval-based assistant for a Serbian higher-education institution, fielding student and staff questions about exam schedules, deadlines, and general academic information. Hybrid retrieval splits factual exam queries to a live MySQL table so answers always reflect the current schedule, and routes everything else to semantic search over scraped institutional content (Pinecone, paraphrase-multilingual-MiniLM-L12-v2 embeddings, Gemini synthesis). Custom ingestion pipeline transliterates Cyrillic to Latin, strips HTML boilerplate, and chunks with RecursiveCharacterTextSplitter before embedding. Conversation history and authentication persisted in MySQL; Streamlit frontend for student-facing access.

Implementation: Pinecone, MySQL, Gemini, Streamlit, Hybrid RAG, paraphrase-multilingual-MiniLM-L12-v2, RecursiveCharacterTextSplitter, Python

AI agent-driven document chunking pipeline that analyzes document structure and contextualizes chunks with logical boundaries. The agent determines optimal split points based on semantic meaning, section hierarchy, and content type. Falls back to LangChain splitters (RecursiveCharacterTextSplitter, MarkdownHeaderTextSplitter) when agent segmentation is not cost-effective. Cleans headers and footers, normalizes metadata for vector storage, detects tables, and generates image descriptions for embedded visuals. Built for heterogeneous, large-scale document collections where standard splitters break: mixed formats, messy structure, tables, visuals, and schema normalization across sources.

Implementation: Python, AI Agents, RecursiveCharacterTextSplitter, MarkdownHeaderTextSplitter, Table Detection, Metadata Normalization, Pinecone

Specialized RAG system for building code and construction regulation documents. Parses legal text with structural awareness - chapters, sections, subsections, and amendments - to maintain hierarchical context. Retrieves exact passages with citation metadata (document, chapter, section, paragraph) so every answer points to the specific regulatory source. Source verification step cross-checks retrieved citations against the original document to prevent hallucinated references.

Implementation: Python, RAG, Citation Engine, Legal Document Parsing, Source Verification, Pinecone

Autonomous crawling agent that structures platform documentation into a queryable knowledge base. Crawls unstructured sources, extracts meaningful content, and organizes it for team and customer access. Generates step-by-step instructions on how to use the web app based on documented features and workflows. Self-updating architecture that detects documentation changes and regenerates affected sections.

Implementation: Python, Web Crawling, Knowledge Base, Content Extraction, Auto-Update Pipeline

Query interface over generated knowledge bases. Delivers instant, specific answers from organizational data - not web search. Retrieves from collected company knowledge with source attribution. Designed for teams that need fast answers without hunting through documentation.

Implementation: Python, RAG, Vector Search, Source Attribution, Next.js

Multimodal extension to the text-based assistant. Users ask questions in natural language and the system retrieves answers directly from a vector store of images - product manuals, installation diagrams, technical sheets - without OCR or caption generation. ColPali and ColQwen encode each image into late-interaction embeddings that capture fine-grained visual and textual features. At query time, the user's question is embedded and matched against this image vector store via similarity search, returning the most visually and semantically relevant pages. Answers are grounded in the retrieved image content combined with text documentation.

Implementation: ColPali, ColQwen, Multimodal RAG, Next.js, Python, Pinecone

Website assistant that answers questions about product catalog, compares multiple products side-by-side, and delivers detailed installation instructions - all in German. Collects contact details through guided flows with suggested questions to reduce friction. Analytics dashboard tracks conversations, user feedback, and conversion paths. FAQ system with confidence scoring to speed up self-service resolution.

Implementation: Next.js, OpenAI, Analytics Dashboard, Lead Capture, FAQ System, Python

Retrieval-based assistant over company documentation - product specs, service descriptions, policies, and FAQs. Two response modes: concise answers for quick lookup, and expanded mode that retrieves additional context from related documents for deep dives. Auto-generated suggested questions based on retrieved context guide users to relevant follow-ups. Analytics track which questions are asked most and where the knowledge base has gaps.

Implementation: Next.js, OpenAI, RAG, Vector Search, Conversation Analytics

AI that answers your phone calls, takes bookings, and handles the full reservation flow through natural conversation. Built on SIP with real-time speech processing, it integrates directly into your dispatch system - no hold times, no extra operators, no missed calls.

Implementation: LiveKit, Google Gemini, Python, Prompt Engineering, Validation Flows, Webhook Orchestration

Conversational AI using Google Places API for venue discovery. Follow-up questioning narrows requirements until search criteria are clear. Web scraping and reranking filter results. Generates personalized outreach emails for selected venues. End-to-end flow from discovery to contact in a single conversation.

Implementation: Google Places API, Python, Web Scraping, Reranking, Email Generation, React

React Native mobile application with AI voice conversations. Vapi.ai handles real-time speech processing. Supabase provides authentication and data persistence. RevenueCat manages In-App Purchase subscriptions and monetization. Production-ready iOS build for App Store deployment with full subscription lifecycle handling.

Implementation: React Native, Expo, Vapi.ai, Supabase, RevenueCat, WebRTC, TypeScript

LangGraph state machine with conditional edges for contextual routing between conversation personas. Real-time speech classification determines which agent handles each user turn. ElevenLabs TTS for natural voice output.

Implementation: LangGraph, LangChain, Google Gemini, ElevenLabs TTS, Speech Recognition, Python

Filtered a large candidate URL set down to usable sites before extraction. Crawl4AI as the primary scraping engine for high-performance, JavaScript-aware crawling at scale. Proxy rotation and browser automation handle difficult, JavaScript-heavy, and protected sites. Vision AI fallback (Gemini) processes sites that block even headless browsers. Pydantic schema enforcement ensures structured, consistent output regardless of source heterogeneity. Extracted pricing tiers, feature lists, and company metadata with precision.

Implementation: Python, Crawl4AI, Gemini Vision, Pydantic, Proxy Rotation, Stealth / Anti-Bot

Natural-language analytics interface combining structured SQL querying with semantic vector search. Routes queries between OpenAI and Google Gemini based on complexity. Supabase integration for data management, authentication, and real-time sync. Automated retrieval with explanation generation and analytics tracking dashboard.

Implementation: Text-to-SQL, OpenAI, Google Gemini, Supabase, Vector Search, Python, Hybrid RAG

Natural language interface for deeply nested tree/folder structured data. Parses hierarchy of machine specifications, maintenance schedules, and operational status stored in folder structures. Plain-language queries return precise answers without manual navigation through nested directories.

Implementation: Python, LangChain, OpenAI, Tree Parsing, Structured Data, FastAPI, React

n8n automation that handles more than 1,000 candidates per run across web search, Google Maps, and app listings. It normalizes and deduplicates records, finds official websites and social profiles, extracts public contact details, verifies email addresses, analyzes fit with OpenAI, drafts personalized outreach, and writes qualified leads to Google Sheets with PENDING_REVIEW status for human approval.

Implementation: n8n, OpenAI, Google Maps API, Google Sheets, Web Scraping, Lead Enrichment, Email Verification

Automated evaluation and testing pipeline for LLM question-answering, retrieval quality, and answer reliability. Built custom metrics alongside DeepEval framework and NLP algorithm metrics for comprehensive benchmarking of retrieval and generation quality. Golden datasets enable rapid comparison of prompting strategies, chunking approaches, and model selections across different providers. Continuous improvement pipeline that feeds evaluation results back into system tuning.

Implementation: Python, DeepEval, Custom Metrics, NLP Metrics, Benchmarking, OpenAI, Anthropic

End-to-end molecular simulation pipeline for precision medicine research. Automated protein-ligand docking with AutoDock Vina and GNINA. RAG system built on research papers and clinical datasets, integrated into CellType CLI for interactive research analysis. Proposed AxialBridge patient app for EULAR - ML/CV-based Axial Spondylitis detection from MRI and ultrasound imaging.

Implementation: AutoDock Vina, GNINA, RAG, ML/CV, MRI Analysis, Python, CellType CLI

Neo4j knowledge graph combined with vector embeddings for hybrid retrieval. Outperforms vector-only approaches on multi-hop queries by following relationship paths. Outperforms graph-only approaches on semantic similarity matching. Best of both architectures for complex reasoning tasks.

Implementation: Neo4j GDS, Vector Search, Hybrid RAG, Knowledge Graph, Python, Embeddings

Automated SFT pipeline via API with pre-flight validation. Training file format checking prevents costly failed jobs. Cost estimation before submission. Real-time monitoring of training progress and evaluation metrics. Reusable system that protects against common fine-tuning mistakes.

Implementation: Python, SFT, OpenAI API, Training Validation, Cost Estimation, Monitoring

MLX-based pipeline for Apple Silicon: LoRA fine-tuning, embeddings generation, reranking, and contrastive learning. Core tasks benchmarked with MTEB for quality validation. High-throughput local training on Apple Silicon with BF16 precision. Self-hosted training without cloud dependency.

Implementation: Python, MLX, LoRA, Apple Silicon, MTEB, BF16, InfoNCE