Bio
Nima Shahbazi is the President of Mindle.ai, who won the biggest computer science or AI competition on record, the $1M Zillow Prize, for creating the most accurate home valuation algorithm. On Kaggle, a Google-owned data science competition platform with 5M members, Nima's models consistently rank in the global top 10. He has won several competitions and achieved Grandmaster status.
His speaking experience includes TMLS for Finance, Kaggle Days in Toronto, ReWork Deep Learning Summit, ReWork Machine Learning Summit, ReWork Deep Learning in Finance Summit, and RBC Disruptors, with 500 attendees and an online audience of more than 174,000.
Nima co-founded Deepnify, an ML SaaS company that worked to reduce food waste. Deepnify raised seed funding and was accepted to the NextAI and Creative Destruction Lab accelerators. He previously worked on big data streams and association rule mining, specifically in capital markets.
AI Awards
Nima was part of the team that won the $1M Zillow Prize for creating the most accurate home valuation algorithm. The algorithm won by a comfortable margin, beating more than 3,800 teams from 91 countries. The team's improvements helped push the Zestimate nationwide error rate from 4.5 percent to below 4 percent across 110 million homes.
Learn more
Nima was part of the prize-winning team that ranked 2nd in the 2018 ACM Recommendation Challenge against over 1,000 data science teams. The algorithm helped Asia's leading music streaming service recommend more relevant music to its listeners.
Learn more
Nima was part of the 2nd-place prize-winning team in the Mercari Kaggle competition with 2,000 competitors. Mercari, Japan's biggest community-powered shopping app, challenged data scientists to build an algorithm that automatically suggests the right product prices to sellers.
Learn more
Nima was part of the 2nd-place prize-winning team in the Two Sigma Financial Modeling Challenge against 2,000 teams. The model searched for signal in financial markets data with limited hardware and computational time, accurately predicting financial movements.
Learn more
Nima ranked 2nd and was a prize winner in the Rossmann Store Sales Kaggle competition against 3,738 data scientists. Rossmann operates more than 3,000 drug stores in 7 European countries. The model forecasted daily sales using historical sales, product, store, weather, and competitor-distance data.
Learn more
Nima was part of the 2nd-place prize-winning team in the Home Depot Product Search Relevance Kaggle competition against more than 2,000 data science teams. The algorithm helped improve customers' shopping experience by accurately predicting search-result relevance.
Learn more
Nima placed 7th against 340 teams in the ICDM 2015 Drawbridge Cross-Device Connections Competition. Given usage data and fabricated non-personally-identifiable IDs, competitors made individual user connections across digital devices.
Learn more
Nima ranked 7th in the Grupo Bimbo Inventory Demand Forecasting Kaggle competition against 1,969 teams. The model forecasted daily sales using historical sales, product, store, and weather data to help Grupo Bimbo maximize sales and minimize returns in more than 1M stores.
Learn more
Nima was part of the team that placed 5th in the Bosch Production Line Performance Competition. Bosch challenged Kagglers to predict internal failures using thousands of measurements and tests made for each component along the assembly line.
Learn moreSpeaking
It takes sizzling passion to win the most competitive data science competitions in the world. Students, event organizers, and attendees can attest that Nima's passion comes across on stage.
Whether speaking to experienced data scientists or non-technical executives, Nima's data-love and obsession captivates audiences, leaving them excited and curious about the future of machine learning.
Nima was selected as a speaker for ReWork 2018 Machine Learning Summit in Montreal, where he presented on demand forecasting.
View source
Nima also spoke at RBC Disruptors with a live audience of 500 and 174,000+ online views.
View source
Invited speaker for Canada AI Day, joining public discussion around AI and trust.
View sourceNews
Coverage from the original speaker page, including Zillow Prize coverage, Deepnify, and Kaggle competition interviews.
Shahbazi has competed in 20 Kaggle events, coming second five times and finishing in the top 15 another five times. He finished in the top five per cent of challengers on contests for everything from Grupo Bimbo inventory to Home Depot search relevance, drug store sales, music recommendation, and the 2017 Data Bowl.
Zillow has slowly improved its Zestimate from a median error rate of 14 per cent when it started in 2006 to 5.7 per cent when the contest began in mid-2017. Once the winners' tweaks are incorporated, the company expects the error rate to dip to about 4 per cent.
Jordan Meyer of the United States, Chahhou Mohamed of Morocco, and Nima Shahbazi of Canada bested more than 3,800 teams representing 91 countries with an algorithm that beat Zillow's benchmark model by approximately 13 percent.
It is amazing to know that millions of people will benefit from our ideas. We brought every novel idea we could to our code and kept experimenting.
On average, Zillow said, the Zestimate is $10,000 off the actual sale price for a median-priced home of about $223,900, and the information gleaned from the Zillow prize winnings could shave $1,300 off that discrepancy.
Deepnify helped food companies predict daily demand for their products and reduce food waste by analyzing a mix of historic sales data and consumer trends.
Nima Shahbazi took second place in the competition, using his background in data mining to gain an edge. By fully exploring and understanding the dataset, Nima was able to engineer features that many participants overlooked.
Asked to search for signal in financial markets data with limited hardware and computational time, this competition attracted over 2,000 competitors.
The winning team leveraged NVIDIA TITAN Xp GPUs for both training and inference, using cuDNN-accelerated Keras and TensorFlow deep learning frameworks.
Shahbazi worked across continents and multiple time zones with a team of two others to beat the Zestimate algorithm. The team's winning solution beat the Zillow Benchmark Model by over 13 percent.
Media
The legacy Tilda page ended with a media section built around this YouTube video.
Start a conversation