Portfolio

Projects & Research

Interesting problems, rigorous solutions — open-source tools and published research you can build on.

Applied MLReranking & Policy OptimizationIn Deployment

governed-rank: Governed Reranking for Any Domain

Steer ranked lists toward any policy objective — fairness, safety, fraud triage, content moderation — without breaking accuracy. Three-step pipeline: orthogonalize, protect, project. 99 KB, minimal dependencies.

0.77→0.92 AIR improvement passing the 4/5ths rule

View case study
Applied ML / Decision TheorySelective Prediction & UncertaintyIn Deployment

Confidence Gate Theorem: When Should Ranked Systems Abstain?

When should a ranked system skip a decision instead of guessing? Two cheaply testable conditions predict whether confidence-based abstention will help or hurt. The key determinant: is your uncertainty from missing data, or from a changing world?

0 violations Every threshold improves accuracy on MIMIC-IV hospital data

View case study
Applied MLAd Tech & PersonalizationIn Deployment

Cookieless Personalization: Session-Level Intent Without Tracking

A complete pipeline for ad and recommendation personalization that uses only session-level signals — no cookies, no persistent user IDs, no cross-site tracking. Three components work together: IntentLens detects what the session wants, the Confidence Gate decides when to trust that detection, and governed-rank steers the ranking without degrading relevance. Validated across 3 public datasets.

4.9x Conversion lift, HIGH vs MEDIUM tier

View case study
Clinical OperationsHealthcare AIIn Deployment

Confidence-Gated Prior Authorization: Automating Healthcare Triage at the Pathway Level

A confidence-gated triage system for prior authorization that detects care pathways from diagnosis and procedure codes, measures confidence in that detection, and routes requests to the appropriate review level — auto-approve, nurse review, or physician review. Validated on 10,000 real MIMIC-IV hospital encounters with 38.3% auto-approve rate and zero monotonicity violations.

38.3% MIMIC-IV encounters routable without human review

View case study
Applied ML / Information Retrieval / Healthcare NLPRetrieval-Augmented Generation & Agentic SystemsExperimental

Graph RAG Without the LLM: Bayesian Community-Biased Retrieval with Calibrated Abstention

Every Graph RAG system today requires an LLM in the retrieval loop — for entity extraction, community summarization, and deciding when to retrieve more. We show that a Bayesian inference pipeline can serve as the entire Graph RAG stack: soft community detection, structured retrieval, and agentic decision-making. On MIMIC-IV clinical data, community-biased retrieval achieves 63.8% precision lift over cosine-only search on adversarial queries, the agentic loop upgrades 87.5% of uncertain cases to high confidence, and the system runs at 2.65ms — not 3 seconds.

+63.8% Precision@5 improvement over pure cosine search on adversarial (conflicting-code) queries

View case study
AI SafetyFairness & GovernanceIn Research

GBP-Audit: Safety-First Bias Correction for AI Models

Safety-first bias correction that knows when not to act. Geometric coherence distinguishes proxy bias from task-aligned bias, five guardrails prevent harmful interventions, and governance packets provide full audit trails. 3 of 7 datasets safely corrected, 4 correctly abstained. Zero accuracy degradation.

7 Real-world datasets across lending, employment, and public policy

View case study

Additional projects in healthcare AI, decentralized systems, and education innovation are moving through the Haske Labs pipeline.

Working on a problem where these tools might help?

Let's talk