Daniel Haskell
Data Scientist
Applied ML projects that show their work: what the model saw, why it decided, and where it fails.
Philemon
Everything I've shipped
Tier 480%less time spent on weekly event setup
ERP / OperationsHestia
Event Booking System
A booking and payments system built for a real client, replacing manual signup tracking over email.
- Full-Stack Delivery
- Payments Integration
- Booking Systems
- Client Discovery
Tier 3$9,350saved per 1,000 loans, vs. a naive 50% cutoff
FintechThemis
Credit Default Risk Model
Calibrated XGBoost tuned on an asymmetric cost matrix, with SHAP explainability and a live FastAPI service.
- MLOps
- Cost-Sensitive Learning
- Model Calibration
- Explainability (SHAP)
Tier 325.9%holdout MAPE forecasting next-day SPY volume
FintechPythia
Trading Volume Forecasting
ETS shipped; XGBoost wins short-horizon validation, then drifts on the holdout. Demo, not a deployed service.
- Time Series Forecasting
- Prophet
- Model Benchmarking
- MLOps
Tier 219/19retrieval hit rate — 23/23 faithfulness and answer relevance
FintechMnemosyne
SEC Filing Q&A
Answers questions about a company's SEC filings and cites the exact paragraph each fact came from.
- RAG
- Vector Search
- Eval Harness
- Claude API
Tier 260%agreement between VADER and DistilBERT across 51,643 real comments
NLP / ResearchEris
Sentiment Model Comparison
Runs two sentiment-reading methods over real Reddit comments and shows exactly where they disagree.
- Fine-Tuned Transformer
- Sentiment Analysis
- Model Comparison
- GPU Inference
Tier 2r = -0.22opening sentiment vs. 5-day return, across 42 real filings
FintechCassandra
Earnings Sentiment vs Stock Movement
Tests whether upbeat language in an earnings release predicts the stock going up afterward. It doesn't, if anything it's mildly contrarian.
- NLP Sentiment Analysis
- Financial Data
- Statistical Testing
- Data Visualization
Tier 2100% client-sidereal job postings scored against a resume, in-browser
HR TechHarmonia
Resume-Job Match Scorer
Scores how well a resume matches a job description, the same way an applicant tracking system does.
- Sentence Embeddings
- Semantic Search
- Cosine Similarity
- On-Device Inference
Tier 2Free text →parsed sets, PR prediction, and a readiness score
Personal ToolMilo
Gym Progression Tracker
Paste a workout and get a training calendar built from it, entirely client-side.
- On-Device Inference
- Time Series Forecasting
- Local-First Storage
Tier 111 rulesvalidated against 60 real SEC-filed contracts
Legal TechHorkos
Contract Risk Flagging
Reads a contract and flags the clauses that are missing, one-sided, or expensive if things go wrong.
- NLP
- Rule-Based Systems
- Risk Scoring
- Regex Parsing
Tier 137real inspection records mapped, Camden County
Public DataHygieia
Health Inspection Tracker
Maps every public health inspection by location, so you can see which restaurants nearby actually passed.
- Geospatial Analysis
- Data Visualization
- Public Records ETL
- Leaflet
Tier 121NJ counties compared, flood risk vs. home price
Public DataDeucalion
Flood Risk vs Home Price
A bivariate choropleth testing whether the housing market actually prices in flood risk.
- Bivariate Choropleth
- Geospatial Analysis
- GeoJSON
- Data Visualization
Tier 1Structured leadscaptured from natural conversation, no forms
Conversational AIPeitho
Lead Qualification Chatbot
A chatbot that qualifies leads by actually talking with them, so only real prospects reach a human.
- Conversational UX
- Lead Scoring
- LLM Orchestration
- Chatbot Design
- Tier 1Building
multi-agent research, both sides of the case
FintechPlutus
Investment Research Agent
Multiple AI analysts argue opposite sides of a stock pick, so you see the real case for and against it.
- Multi-Agent Systems
- LLM Orchestration
- Cost-Optimized Inference
- Financial Analysis
About
I'm a machine learning engineer and data scientist. B.S. Data Science, Minor in Computer Science, Rutgers University.
I build systems end-to-end: framing the problem, training and validating the model, and shipping it behind a live API.
Education
B.S. Data Science, Minor in Computer Science
Rutgers University · June 2026
- Machine Learning
- Data Structures & Algorithms
- Statistical Modeling
- Regression Methods
- Q-Learning
- Applied ML Concepts
Security
Yearly CTFs
- DEFCON Qualifier
- Texas Security Week
Contact
- Emaildjhask@comcast.net
- GitHubgithub.com/bigrolla
- LinkedInlinkedin.com/in/danielhaskell2
