Goldsmith & Co. AI Lead & Software Engineer

Jan 2025 — present · New York, NY
  • Sole engineer at an executive search firm: I own every production system end to end, including a Rails and Postgres API, a client-facing portal (JavaScript, OAuth 2.0 with PKCE, role-scoped authorization), a Snowflake reporting layer, and an internal AI platform, and I set technical direction with firm leadership.
  • Built the human correction loop for a production data pipeline: recruiters edit records inline in the portal, corrections land in an override store, an admin is pinged for review, and the accepted correction feeds back into the system of record. Real annotators who are not engineers, on data the business depends on.
  • Designed and shipped an agent platform (LangGraph.js and the Claude API) now in daily use by non-technical staff, plus the evaluation framework measuring whether its output is actually correct, including adversarial testing against a curated attack corpus.
  • Kept that portal healthy while requirements shifted weekly: automated build and deploy, render-escaping invariants, contract tests, and a full security-audit remediation round.
  • Diagnosed a silent 8% data loss in a production sync pipeline (4,840+ failed syncs, missing history rows, phantom records) that had corrupted downstream reporting undetected. Traced root cause, repaired the historical record, and added invariant checks so the failure surfaces immediately rather than silently.
  • Supervised an ML intern on a learning-to-rank project (LambdaMART, NDCG). Established by measurement that ranking quality was structurally capped by label sparsity rather than model choice, and redirected the work instead of tuning against a ceiling. The labels were the bottleneck, not the model.
railspostgreslanggraph.jsclaude apisnowflakeoauth 2.0 / pkcejavascripteval design

Athena Founder & Engineer

2025 — present · New York, NY
  • Founded a consumer-protection startup detecting synthetic media and non-consensual imagery. Applied to Y Combinator.
  • Trained and deployed an image classifier (EfficientNet-B0, PyTorch) with perceptual hashing for near-duplicate matching: 0.996 accuracy on a held-out set of real and synthetic images.
  • Set decision thresholds against the asymmetric cost of a false positive versus a false negative rather than optimizing a symmetric metric, and maintained an eval set built to detect drift as the underlying generative models moved.
pytorchefficientnet-b0perceptual hashingthreshold calibrationmodel cards

CLEAR Data Analyst

Jan 2024 — Jan 2025 · New York, NY
  • Built churn forecasting models across a 20M+ member base at a biometric identity company (NYSE: YOU), and ran scenario analysis on how pricing changes traded retained volume against revenue.
  • Automated a $1B+ revenue forecast, replacing a hand-assembled analysis with a reproducible pipeline that could be re-run and checked against realized results.
snowflakesqlpythonforecastingscenario analysis

National Museum of Mathematics Integrator

Jan 2025 — present · New York, NY
  • Explain mathematical ideas on the museum floor to visitors of every age, adjusting to the person in real time. The best training I have had in finding the explanation that actually lands for the person in front of me.
expositionmathematics

Estée Lauder Companies Software Development Intern

Jun 2022 — Aug 2022 · New York, NY
  • Designed and built an internal IT web application in Python (Flask, SQLAlchemy) with user authentication and secure access controls.
  • Built modular APIs and data-processing scripts to automate report generation, delivered on a two-week sprint cadence.
pythonflasksqlalchemysql
Education

University of Michigan. B.S. Pure Mathematics; Studies in Ukrainian.

University of Michigan

Aug 2020 — Dec 2024 · Ann Arbor, MI
  • James B. Angell Scholar (2021, 2022), awarded for consecutive terms of all-A academic standing.
  • Selected by the Department of Mathematics to deliver the Departmental Commencement Address, Class of 2024.
  • Officer, Women in Mathematics. ICPC-style competitive programming.
  • Coursework across machine learning, artificial intelligence, database systems, operating systems, software engineering, and data structures and algorithms, on a pure-mathematics core of topology, analysis, and algebra.
Technical

Grouped by what I use it for rather than by how long the list can be made.

The toolchain. Everything here appears in a repository or a production system.

AI systems

  • Claude API
  • LangGraph orchestration
  • MCP servers and clients
  • SFT / DPO on preference data
  • RAG with cited retrieval
  • Human-in-the-loop approval gates

Evaluation

  • Eval harness design
  • Grader calibration
  • Held-out set construction
  • Adversarial / red-team corpora
  • Annotation schema design
  • Quality gates in CI

Languages

  • TypeScript / JavaScript
  • Python
  • Ruby
  • SQL
  • Go
  • C++

Backend & data

  • Rails, Sinatra
  • PostgreSQL, Snowflake
  • Kafka, Flink, Iceberg
  • TimescaleDB
  • pandas, NumPy, scikit-learn
  • PyTorch

Infrastructure

  • Docker
  • Terraform, Helm, Kubernetes
  • GitHub Actions CI/CD
  • AWS, Heroku, Render
  • Linux
  • Air-gapped delivery

Front end

  • HTML / CSS / JS
  • React, Angular
  • esbuild
  • OAuth 2.0 / PKCE
  • Review interfaces for non-technical users

Mathematics

  • Topology
  • Stochastic calculus
  • Graph theory
  • Linear algebra
  • Probability and statistics
Next

The project list is where the detail lives. Every claim there points at running code.