AI systems builder focused on turning workflow friction into production-grade AI applications.
I build LLM-driven systems, analytics products, and decision tools across product, backend, and evaluation workflows.
- AI Systems & Eval Platforms: multi-model orchestration, prompt pipelines, judging workflows, human-in-the-loop review
- Analytics & Decision Tools: dashboards, KPI systems, forecasting workflows, business intelligence apps
- Productized Internal Tools: operational consoles, workflow automation, data products, scoring systems
- End-to-End Prototypes: fast MVPs using modern full-stack tools with real backend logic
Languages: Python, SQL, JavaScript, R
AI / Data: PyTorch, pandas, scikit-learn, LLM orchestration, evaluation systems, experimentation, RAG, feature engineering, feature stores
Big Data / ML Systems: PySpark, Apache Spark, Kafka, Flink, Databricks, Snowflake, BigQuery, Delta Lake, dbt, Feast, MLflow
Backend / Infra: Supabase, Postgres, REST APIs, Git, Docker, Kubernetes, Google Cloud
AI Dev Tools: Cursor, Claude Code, Codex, Google Antigravity, Lovable
Analytics / BI: Power BI, Tableau, ETL, data modeling
Other: Streamlit, Jupyter, RStudio
| Project | What It Does |
|---|---|
| AI Evals Orchestration | End-to-end evaluation workflow for comparing model outputs, saving decisions, and supporting human review |
| Trust Pulse | Subscription abuse operations console linking accounts, payment methods, devices, IPs, and clusters for investigation |
| Recommendation Quality Lab | Offline recommender experimentation lab with temporal splits, candidate generation, ranking, and beyond-accuracy metrics |
| Behavioral Prediction Engine | Hackathon group project combining behavioral profiling, ML, and LLM-based prediction to forecast how real people would answer hidden survey questions |
| Context Parking | Structured context management system for saving, retrieving, and reusing work context across AI workflows |
| Power BI Apps Showcase | Showcase of 15 Power BI apps built for SMB reporting across sales, finance, inventory, operations, and production |
- Building AI products that look and feel production-ready
- Combining product thinking with backend execution
- Exploring eval systems, trust/risk workflows, and decision-support tools
- Shipping projects that are recruiter-friendly and demoable
- LinkedIn: https://www.linkedin.com/in/jasjyot-singh-14a8aa217/
- Email: jasjyotsingh.work@gmail.com
- Phone: +1 (951) 227-0134
Full-time opportunities in AI Product Management, Product, Data, and analytics-focused roles across the US, especially San Francisco and other major cities.