Jan 2026 – Present · 5 mos

• Chat-first AI analyst for regulated enterprises, with a deterministic execution engine designed for audit (coherany.ai)
• 400k decisions/sec at 12ms p95: architected a two-layer system separating conversational discovery from production execution. An LLM chat layer interprets analyst questions; the execution layer matches records via FAISS HNSW with no LLM calls in the hot path, so every classification is reproducible byte-for-byte under audit.
• Built enterprise multi-tenancy end-to-end: PostgreSQL row-level security, four-role RBAC,
• per-tenant quotas, JWT with default-deny middleware, Redis-backed distributed rate limiting, zero-downtime API key rotation. SOC 2 in progress with active Fortune 500 design-partner conversations.
• Delivered 270+ REST endpoints and eight ingest adapters (CSV, PDF, audio, video,
• PostgreSQL, REST, S3, SFTP) with Prometheus observability and hash-verified audit logging.
• Built a visual pipeline Studio (144 nodes, 1,507 templates) so non-technical operators such as VP Claims, Chief Compliance, and Chief Fraud can configure discovery and approval flows without writing Python.
• Validated predictive performance on public datasets: NASA C-MAPSS turbofan (3.3× anomaly-detection premium), Woolsey wildfire / NOAA (8-day early signal), CMS Dialysis Care Compare (triaged 7,557 facilities in under one analyst-day), Medicare DMEPOS fraud (3.0–6.6× detection uplift). 10,200+ automated tests, circuit breakers, exponentialbackoff recovery.

Oct 2022 – Dec 2025 · 3 yrs 2 mos

• Production systems and architecture for Fortune 50 clients in healthcare, insurance, telecom, and digital payments
• Owned a $2M ARR portfolio of eight Fortune 50 clients averaging $4B in annual revenue, leading projects from initial architecture through production cutover.
• Architected a financial data platform processing ~20M transactions/day for a digital payments company. Delivered in three months with full auditability and compliance sign-off.
• Engineered a financial processing solution that cut go-live time by 50% and generated $750K in recurring revenue in its first year.
• Replaced a Fortune 50 client's manual reference-data SOPs with an Azure Data Factory pipeline and Power Apps front end. Design, build, and deployment owned end-to-end.
• Built automation frameworks that cut deployment cycles 5×. Established an engineering enablement program across 30+ team members; onboarding time fell 60%.

Mar 2017 – Oct 2022 · 5 yrs 7 mos

• Owned platform upgrades with QA gates that reduced production incidents from 2–3 hotfixes per release to zero across the next 14 releases. Built event-based accounting models and automated testing solutions in Gradle and Java for high-volume insurance data processing.
• Delivered Multi-GAAP, US GAAP, and IFRS-compliant financial software implementations end-to-end. Improved ETL performance on PostgreSQL, MS SQL, and Oracle.

Jan 2014 – Jan 2017 · 3 yrs · Boston, MA

• - Led the enterprise ICD-9 → ICD-10 transition under HIPAA, designing code mappings and validation rules. Built SQL pipelines on MS SQL Server, SSIS, and SSRS for finance, actuarial, and claims teams.

Jan 2013 – Jan 2014 · 1 yr · Greater Boston

• - Owned analysis, design, testing, and production support for SQL Server systems serving high-volume automotive retail clients.

Jan 2012 – Jan 2012 · 0 mo · Singapore

• - Built an agent-based model of endocytic trafficking from multi-lab datasets to reveal how specific genes influence vesicle movement.