Tyler Sacharow

Coursework: Honors Linear Algebra; Honors Multivariable Calculus; Data Structures and Algorithms; Computer Systems; Engineering Product Management

• Built the entire technical stack behind a decision research firm that tells companies what their customers will do before the money is spent: a simulation engine that builds populations from real interviews, a self-improving model grounded in published human-simulation research, and the pipeline and internal tools behind every signed briefing
• Held to a public record no one else in the category publishes: 390 questions from 24 national surveys across 9 research organizations, every one answered before the results were seen and each prediction shown next to the real number

• Co-founded and led all engineering for a content protection product that shields web content from AI scrapers while keeping it fully readable to humans; built the core engine and the production system
• Won Penn's $75,000 Perlman Grand Prize, first of 100+ ventures in the Venture Lab Startup Challenge

• Shipped production features across a full-stack TypeScript monorepo (Next.js, Node.js, Turborepo, PostgreSQL) for a firm with a 100% success rate across 400+ O-1A cases, serving founders and enterprise employers
• Architected AI-driven data pipelines that process visa petitions and validate structured LLM outputs, cutting attorney case preparation from the industry's 6–8 weeks to under 2
• Built end-to-end case management workflows handling O-1A, EB-1A, and EB-2 NIW visa applications with Doppler secrets management, pnpm workspaces, and automated CI/CD pipelines

• Built Penn Marketplace, a student marketplace for items and sublets serving 10,000+ users, with Next.js, Django REST Framework, PostgreSQL, and Redis
• Engineered a polymorphic listing API supporting 2 listing types through unified endpoints with 14 filterable fields, 8 database indexes, and N+1 query elimination
• Implemented Twilio SMS phone verification with Redis-cached codes gating an offer negotiation system with database-enforced spam prevention and automated content moderation

• Organizing exclusive VC firm visits (Notable Capital, Inspired Capital) in NYC for 25+ top student founders from Harvard, Princeton, Yale, Columbia, Penn, and other Ivy League universities
• Managing operations for an intercollegiate 501(c)(3) network connecting 120+ student entrepreneurs across technology, investing, and venture capital

• Co-founded an international competition reaching 350+ participants from 40+ countries, lowering entry barriers to advanced CS education
• Led problem development, outreach, and logistics for the Massachusetts Computer Science Olympiad
• Managed $50,000+ prize pool; secured sponsors (Jane Street, Google, JetBrains)

• Grand Prize Winner (3rd place overall) at HackMIT 2026 and 1st place in the OpenAI, ASUS, and Dimensional sponsor tracks (~$15,000 in prizes); built in 24 hours at MIT
• Built a wheelchair-mounted robotic caregiver that bathes, feeds, medicates, and monitors vitals for elderly and disabled users: two 6-axis OpenYAM robotic arms, an Intel RealSense D455 depth camera with MediaPipe pose estimation and RANSAC body reconstruction, Arduino sensing, and a Deepgram Nova-3/Aura-2 voice agent whose every command is validated before it reaches the arms
• Owned the real-time 3D layer in Three.js: a rigged avatar that mirrors the user's live pose, renders the arms' planned paths, and animates bathing, feeding, and medication routines as each body region completes; integrated the team's perception, robotics, hardware, and AI agent into one working system

• Won the $25,000 Grand Prize (1st place overall) at DNHacks 2026 in Washington, D.C., a 200-builder invitational for Ivy League, MIT, Stanford, Caltech, Berkeley, and UChicago students backed by Second Front, Clearview AI, Craft Ventures, DTX Ventures, and OpenAI
• Built a real-time structural health console for carbon-fiber concrete that senses its own load: one piezoresistive physics model driving a 62 m LiDAR-scanned bridge (12.3M survey points, 28 members), a photogrammetry-captured dam (32 monoliths), and a 438-member datacenter with load-path propagation
• Rendered 2.6M-point clouds in one instanced Three.js draw call at 120 fps with 0 dropped frames, streamed 10 kHz live sensor data from a Digilent Analog Discovery into the browser over WebSocket, and shipped thermal compensation, peer cross-checks, and days-to-limit forecasting; 117 Playwright checks, 0 WCAG 2.1 AA violations

• 2x Finalist at YHacks 2026 (Yale): top 10 in the Entertainment and Snap Lens Studio tracks (600+ participants)
• Created a full-stack social commerce platform in Next.js and TypeScript with a low-latency infinite-scroll UI/UX feed, AR avatar experience, and a social purchase system
• Created a client-side recommendation algorithm, converting each product into a 32-dimensional vector, using cosine similarity retrieval, reranking with 5 different signals, and L2 normalization

• Sponsorship Winner at NexHacks 2026 (CMU), the largest hackathon on the East Coast (1,500+ participants)
• Built a Chrome extension that scrapes and synthesizes real user opinions from Reddit, YouTube, and expert review sites to generate product scores
• Created a personalized recommendation engine that ranks products based on aggregated community sentiment and forum discussions