THE ENGINEERING INDEX

Systems,
from the inside.

Command analysis, local browser Q&A, Kafka ingestion, retrieval, and offline messaging. Explore the architecture, interfaces, and tradeoffs behind each implementation.

16 projects found

INTENTGUARD
EXECUTION CONTEXT → RISK ANALYSIS → YOUR DECISION

AI / Command-line risk analysis

IntentGuard

A Python command-review CLI combining shell rules, classifier risk, BM25-ranked command history, and LiteLLM into allow, warn, or block decisions.

PythonLiteLLMBM25SQLiteTyper
  1. Command + user
  2. Rules + classifier
  3. BM25 history + LLM
  4. Risk → execution decision
WEBLMS
DOM + OCR → LOCAL CONTEXT → EXTRACTIVE Q&A

AI / Local browser Q&A

WebLMS

A TypeScript extension that extracts DOM text and tables, runs Tesseract.js OCR, caches page context in IndexedDB, and answers with extractive sentence scoring.

TypeScriptWebExtensionsTesseract.jsIndexedDBComlink
  1. DOM + images
  2. Text + table extraction
  3. Question classification
  4. Extractive answer
RAG AGENT
DOCUMENTS → RETRIEVAL → GROUNDED ANSWERS

AI / Document retrieval & reasoning

RAG Agent

A Streamlit assistant with 1,000-character document chunks, MiniLM embeddings, persistent Chroma cosine search, selected-document retrieval, and optional web-search augmentation.

PythonStreamlitChromaMiniLMLangChain
  1. Files → text
  2. Chunks → MiniLM
  3. Chroma retrieval
  4. Guardrails + answer
APPTRACK
SDK EVENTS → KAFKA → CLICKHOUSE

Systems / Event ingestion & analytics

AppTrack

Go HTTP/UDP collectors accept structured SDK events, publish to Kafka, and feed a ClickHouse consumer for column-oriented event analytics.

GoKafkaClickHouseHTTP / UDP
  1. JS / Go / Python SDK
  2. Go collector
  3. Kafka topic
  4. ClickHouse events
VISION AI
IMAGES + DOCUMENTS → CONTEXT

AI / Document-grounded conversation

Vision AI

A FastAPI and Next.js prototype combining layered PDF extraction, OCR fallback, local MiniLM embeddings, NumPy cosine retrieval, and source-aware document chat.

FastAPINext.jsNumPyMiniLMOCR
  1. Upload + extraction
  2. Text + OCR fallback
  3. MiniLM + NumPy index
  4. Context → chat
NAVIGATORPROMPT
PROMPTS → MODELS → EVALUATION

AI / Prompt orchestration

NavigatorPrompt

A Next.js prompt workspace with FastAPI task endpoints, Celery generation/evaluation workers, Redis result handling, and PostgreSQL-backed prompt history and libraries.

Next.jsFastAPICeleryRedisPostgreSQL
  1. Next.js action
  2. FastAPI task API
  3. Celery + Redis
  4. Status + prompt library
INCIDENT PDF INGESTOR
INCIDENT PDF → ROW EXTRACTION → SQLITE

Data / PDF-to-SQLite ingestion

Incident PDF Ingestor

Downloads public incident PDFs, reconstructs wrapped records with PyPDF2 and regular expressions, inserts normalized SQLite rows, and reports grouped incident counts.

PythonPyPDF2RegexSQLite
  1. Incident PDF
  2. PyPDF2 + row parser
  3. SQLite records
  4. GROUP BY nature
INCIDENT VISUALIZER
SQLITE → GEOCODING → CLUSTERS + CHARTS

Data / Geocoding & incident analysis

Incident Visualizer

A Flask interface that reads incident records with pandas, geocodes addresses, clusters coordinates with KMeans, and renders Matplotlib/Seaborn chart artifacts.

FlaskpandasKMeansOpenCageMatplotlib
  1. SQLite incidents
  2. pandas + geocoding
  3. KMeans + aggregation
  4. Flask chart views
FBI DATA PIPELINE
EXTRACT → TRANSFORM → LOAD

Data / FBI Wanted API normalization

FBI Data Pipeline

A Python CLI that reads FBI Wanted API pages or local JSON, normalizes multi-value fields, and emits a documented three-field delimited format.

PythonRequestsJSONCLI
  1. API page / JSON file
  2. Record normalization
  3. Multi-value fields
  4. Delimited output
RESUME RANKER
PARSE → MATCH → REFINE

AI / Semantic & lexical matching

Resume Ranker

A Flask NLP workflow combining resume extraction, SentenceTransformer embeddings, TF-IDF similarity, and structured skills to rank resumes against job descriptions.

FlaskSentenceTransformersTF-IDFspaCy
  1. PDF / DOCX
  2. Text + NLP features
  3. Semantic + lexical similarity
  4. Ranked result
REDACTED NAME PREDICTOR
REDACTED CONTEXT → FEATURES → NAME PREDICTION

AI / Supervised name reconstruction

Redacted Name Predictor

An NLP experiment using redaction length, neighboring words, POS tags, and n-grams with CountVectorizer and class-balanced logistic regression to reconstruct names.

NLTKscikit-learnLogistic regressionFeature engineering
  1. Redacted TSV contexts
  2. NLTK + feature strings
  3. CountVectorizer
  4. Logistic regression names
UNIVERSITY JOB SCRAPER
DISCOVER → COLLECT → STRUCTURE

Data / HTML extraction & email reports

University Job Scraper

A Requests/BeautifulSoup script that pairs UF job-listing rows with summaries, normalizes listing fields, and composes an SMTP email report.

PythonBeautifulSoupRequestsSMTP
  1. UF jobs HTML
  2. BeautifulSoup rows
  3. Normalized listings
  4. Email report
WEBCHAT
MESSAGES → DELIVERY → CONVERSATION

Systems / Offline-capable messaging

WebChat

React and Dexie/IndexedDB track pending messages; authenticated FastAPI WebSockets deliver room events, while a sync API reconciles offline records and server IDs.

ReactFastAPIWebSocketsDexieSQLAlchemy
  1. React + local pending record
  2. Authenticated WebSocket
  3. Message persistence + room fanout
  4. Reconnect + sync
PERSONAL WEBSITE BACKEND
REQUESTS → SERVICES → CONTENT

Systems / Archived portfolio content API

Personal Website Backend

The original FastAPI service for this portfolio’s projects, journal entries, and contact workflows. The website now uses repository-managed JSON content.

FastAPIRESTContent
  1. FastAPI routes
  2. SQLAlchemy sessions
  3. Azure SQL models
  4. JSON API responses
GEOGUARD
LOCATION → GEOSPATIAL CONTEXT

Systems / History-aware location privacy

GeoGuard

Clusters location history with DBSCAN, identifies frequently visited places, perturbs sensitive coordinates, and integrates with a browser geolocation override.

PythonDBSCANNumPyJavaScriptGeolocation
  1. Location history
  2. DBSCAN + POI detection
  3. Coordinate perturbation
  4. Geolocation override
COMPILER DESIGN
SOURCE → PARSE TREE → VISITOR + SCOPES

Systems / ANTLR parse-tree interpreter

Compiler Design

A Delphi/Pascal grammar and Java parse-tree visitor with nested scope maps, class definitions, object fields, and method metadata.

JavaANTLRParse treesRuntime scopes
  1. Delphi source
  2. ANTLR lexer + parser
  3. Parse-tree visitor
  4. Scopes + object state