Austin French

Role eliminated in company-wide restructuring

AI consultancy embedding engineers directly with clients. Lead engineer on three simultaneous engagements, from technical discovery through architecture, delivery, and executive demos.

• Built retail placement analytics across 1,500+ stores and several million historical sales records (TypeScript/Next.js), with gondola-level heatmaps that let account execs compare store, field vs. non-field, and placement performance. In use by the client; on a held-out historical placement change, the model matched the observed direction of sales movement for 96% of 300+ SKUs.
• Deployed a TypeScript/Next.js mobile scheduling system that replaced WhatsApp-based availability collection and manual schedule assembly, generating hundreds of monthly shifts for 120 representatives across 425 stores
• Built an OCR and two-stage LLM pipeline (Google Document AI + LLM extraction) processing hundreds of printed and handwritten forms weekly, enabling the client to wind down an outsourced data-entry workflow; wrote evals and autonomy gates so over half of documents ran without human review
• Built internal tooling used across engagements: a parallel Claude Code session orchestrator, Slack-to-PR bug automation, and git provenance tracking

Recurring seasonal contract for seven years, concurrent with the full-time roles below

Talent logistics for major live productions: onboarding, scheduling, payments, and compliance for 10,000+ performers across the Super Bowl, Oscars, and Grammys.

• Sole on-site engineer at Super Bowl LX: ran talent check-in on the platform, caught workflow friction live, and turned operational needs into same-day technical decisions
• Cut page loads from 10+ seconds to under 400ms by rewriting SQL and ActiveRecord usage, reducing 500+ queries per page to ~10, under live-event deadlines
• Paid 100% of the halftime show’s non-union performers through the platform via Zeal; integrated Twilio, Yardstik background checks, and NFL SEMS APIs for onboarding and event compliance
• Processed $2MM+ in donations for Boys & Girls Clubs of America’s 2021 fundraiser (Node.js, Express, Stripe)

#1 photo documentation app in construction, used by 140,000+ contractors.

• Led the Pages AI Assistant from ideation to production: contractors speak or type on a job site and get finished documentation instantly. Scaled to ~5,000 requests/day; customers reported saving 2–5 hours per week
• Built shared Rails abstractions on top of RubyLLM that were adopted across the product suite; shipped voice commands, context-aware invoice generation, and tool-calling workflows on Anthropic, OpenAI, and Groq
• Architected Share Link (Rails, React/TypeScript) so users could securely share any asset with clients and payers off-platform
• Cut PDF export size 10x with streaming, compression, and Sidekiq/S3 async processing for 10,000+ daily exports

Industry standard for electronic prior authorization, connected to 950,000+ providers, 50,000+ pharmacies, and nearly every US health plan.

• Launched nine pharmaceutical brands representing $10M+ ARR, expanding supported platform coverage by 40%
• Built Configuration Station (Rails, React, PostgreSQL), reducing technical brand onboarding from months to hours and enabling nontechnical staff to configure launches without engineering
• Technical lead for the Spravato, Renflexis, and Ontruzant launches, owning implementation planning and coordination across engineering and launch stakeholders
• Created VEST, a technical-debt prioritization framework adopted across engineering; mentored 5 developers, including transitioning a CSM into engineering

• Cut a daily pipeline processing 25,000+ customer records from six hours to 45 minutes through batch processing and database partitioning