Scalar Research is an artificial intelligence and data science consulting firm.
We help companies tackle complex business challenges with data-driven products and solutions leveraging cutting-edge machine learning and advanced analytics.
Increase revenue and cut costs by adopting data-driven approaches to optimizing your business.
Price & Inventory Optimization →
Built demand forecasting and pricing models to help companies like ISC (Two Sigma-backed insurance MGA) and a mid-size US retailer optimize profit and inventory.
Operational Efficiency →
Helped a vendor optimize a Fortune 500 auto company's servicing efficiency with computer vision.
Built tooling for a data science team deploying supply-chain analytics in a major CPG conglomerate in LATAM.
Preventative Risk Management →
Advised a global consulting firm and a major Asian industrial conglomerate in the R&D of a computer vision product for industrial safety.
Investment & Financial Services
Gain an edge by applying machine learning on proprietary and public data to drive actionable insights.
Satellite Imagery →
Leveraged large-scale geospatial imagery data to help a reinsurer model wildfire risk in California, and an EU-backed social impact organization detect illegal deforestation in Latin America.
Fraud & Risk Intelligence →
Developed models to detect ACH fraud and estimate transaction risk for a fintech company backed by a major UK bank.
Alternative Data →
Helped a US family office and a major sovereign wealth fund gain investment insights from proprietary data sources with deep learning and natural language processing.
Technology & Startups
Join past clients ranging from seed-stage YCombinator startups to breakout companies like Fandom and Scale AI in developing innovative ML-powered products.
Recommendations & Personalization →
Built personalization and recommendation models for clients such as Fandom and Screenshop, helping their users find relevant content among a catalogue of millions of items.
Ad Targeting →
Implemented multiple large-scale NLP models that enabled Fandom to better understand its users (e.g. interests) and content (e.g. automatically inferring topics, tags, and similar pages) for better ad targeting.
Visual & Semantic Search →
Implemented computer vision models for visual product search for Screenshop, a fashion discovery and shopping app.
Built a semantic search engine for enterprise data for TouchCast using deep learning.
Enriched metadata of massive stock photo collection using computer vision and face processing for a $XB+ media company.
Data Vendors →
Helped two data vendors enhance their offerings by parsing unstructured data (scraped data, PDFs).
Other Projects →
Other projects include:
- Biomedical data mining to predict cancer therapeutics for a YC-backed biotech startup
- Cheating detection for an online gaming company
- Matching engine for a dating app
- Sports tracking and scoring from real-time video
- Automatic speech recognition with deep learning
We've worked with clients ranging from government-backed organizations to innovative early stage startups.
Think your company might benefit from data-driven solutions? Let's talk!
Gabriel Bianconi — Founder
Gabriel is a senior machine learning engineer and data scientist helping companies tackle business challenges with predictive analytics, computer vision, and natural language processing. Past clients include startups backed by YCombinator and leading VCs (e.g. Scale AI, Fandom), investment firms and their portfolio companies (e.g. a Two Sigma-backed insurance MGA), and large enterprises (e.g. an industrial conglomerate in Asia, a leading strategy consulting firm). Some of the machine learning solutions he's built have helped clients close Fortune 500 deals, multiple funding rounds, and an acquisition by a public tech company.
Beyond consulting, Gabriel is a frequent speaker at major technology conferences (e.g. AWS re:Invent 2019, Amazon's annual conference) and other leading organizations (e.g. Tate Modern). He's also a reviewer for top academic conferences (e.g. awarded Top Reviewer at ICML 2020) and AI textbooks (e.g. Packt's upcoming book on deep learning for NLP).
Previously, he received B.S. & M.S. degrees in computer science from Stanford University, where he conducted award-winning research in machine learning. He graduated with honors for his thesis investigating quantum deep learning algorithms using NASA's D-Wave quantum computer, and was selected to present his work at the AQC 2017 Conference in Tokyo. Later, his research focused on improving clinical care and reducing monitoring costs in hospitals by leveraging deep learning and computer vision, resulting in a first-author manuscript selected as Top 10 Research Paper at the NIPS Machine Learning for Health 2017 Workshop, as well as a publication in Nature (npj) Digital Medicine.
HQ New York City