Tome 2.0

• Founding engineer of Tome's new product — an AI CRM / research intelligence tool that provides a single pane of glass into all of your customer relationships. Working up and down the stack (infra, full-stack, AI engineering) to build the first primitives of our beta product.

• Built Tome's foundational contact recommendation system from scratch, processing 400M+ profiles to automatically identify ideal customer champions at target accounts.

• Designed LLM-based contact quality evaluation loop to improve precision: 61% -> 97% and recall: 29% -> 60%, enabling customers to successfully identify five relevant decision-makers per target account

Tome 1.0

• Engineered distributed Redis-backed job queue processing 50K daily requests. This piece of infrastructure enabled six new product features in 4 months and powers Tome's flagship research brief generation product. Implemented logging, autoscaling, testing, and monitoring.

• Architected Typesense-based search cluster and scaled up to 100K collections and 10K daily queries. Drove decisions on search provider selection, privacy considerations, indexing strategies, and query optimization.

• Executed SOC II compliance initiatives: built permanent data deletion pipeline, restricted public media access, conducted firewall review. Enhanced overall system security and data protection.

Worked on various infrastructure teams, broadly helping Airtable scale and maintain its fleet of services and keep those services resilient to internal and external outages.

Most notably, I helped eliminate the biggest single point of failure in Airtable’s architecture (a highly centralized MySQL database). We improved site-wide availability from 0% (SEV-0) to >99% (SEV-3) during outages to this database and prevented five hrs of site downtime in six months.

I also helped create Airtable's inaugural new grad program, leading to our first new grad class of twelve engineers in 2022. I designed interview questions, created onboarding sessions from scratch, and hosted weekly office hours.

Teaching CS170 (Efficient Algorithms & Intractable Problems). Teaching weekly sessions, holding office hours, and designing course content.

Developing planning models for training RL algorithms predicting demand response in an energy-saving, social game framework.