# Mindle.ai > Mindle.ai builds elite applied AI, quantitative finance machine learning workflows, applied prediction systems, and AI products for teams that need models to survive real data, real decisions, and real evaluation pressure. Mindle is led by Nima Shahbazi, an award-winning AI scientist, entrepreneur, Kaggle Grandmaster, Zillow Prize winner, and public speaker at AI and applied learning venues. ## Credits Mindle.ai is led by Nima Shahbazi. Nima Shahbazi credits: - Founder and lead AI scientist behind Mindle.ai. - Entrepreneur and builder of applied AI products, including Olva. - Kaggle Grandmaster with a global top-tier competition record across ranking, forecasting, valuation, recommendation, search relevance, and quantitative modeling. - Winner of the $1M Zillow Prize for automated home valuation, competing against 3,800+ teams from 91 countries. - First inventor on a Zillow patent tied to core Zestimate modeling. Source: https://scholar.google.com/citations?view_op=view_citation&hl=en&user=tHeilS8AAAAJ&cstart=20&pagesize=80&citation_for_view=tHeilS8AAAAJ:3fE2CSJIrl8C - Ranked #1 in Canada and top 20 worldwide on Kaggle, with 10 gold medals and 10 silver medals. - Prize-winning competitor in Two Sigma Financial Modeling, Optiver market prediction, WSDM / KKBox recommendation, Home Depot search relevance, and Rossmann sales forecasting. - Public speaker at top-tier AI and applied learning venues including KDD / Canada AI Day, AIC Ontario, Georgian programs, ReWork, and RBC Disruptors. ## Site - Home: https://www.mindle.ai/ - Services: https://www.mindle.ai/services - Products: https://www.mindle.ai/products - Proof: https://www.mindle.ai/#proof - Contact: https://www.mindle.ai/contact - Email: sales@mindle.ai - Address: Unit 2020, 2 Bloor St East, Toronto, ON M4W 3E2, Canada ## Primary Positioning Mindle builds high-accuracy machine learning systems for prediction, markets, and real-world products. The company works on applied AI systems, hedge fund and quantitative finance ML, product discovery, model evaluation, and AI product development. Core themes: - Elite applied AI for prediction, markets, and real-world products. - Machine learning systems where accuracy, evidence, and evaluation pressure matter. - Competition-grade modeling discipline applied to business, finance, product, and operational problems. - Practical AI product development, including Olva. ## Services ### Hedge fund and quantitative finance ML Mindle supports alpha research, market prediction, signal discovery, feature engineering, model validation, and production research workflows. Focus areas: - Market data modeling - Signal and feature audits - Competition-grade validation ### Applied prediction systems Mindle designs and improves models for valuation, forecasting, ranking, recommendation, search relevance, pricing, and operational decision systems. Focus areas: - Forecasting and valuation - Ranking and recommendation - Pricing and demand models ### AI product discovery Mindle turns ambiguous business problems into low-risk AI product proposals with feasibility, data needs, roadmap, timeline, and cost clarity. Focus areas: - Opportunity workshops - Feasibility memos - Prototype roadmaps ### Model performance rescue Mindle diagnoses underperforming ML systems, repairs evaluation loops, improves benchmarks, and reduces model error with disciplined experimentation. Focus areas: - Benchmark design - Error analysis - Optimization roadmap ### In-house ML capability Mindle helps teams build the data, hiring, evaluation, and process foundations needed to ship reliable ML systems internally. Focus areas: - Hiring support - Data pipeline readiness - Evaluation culture ## Engagement Model 1. Intro and fit: understand the business goal, constraints, and whether Mindle is the right technical partner. 2. Problem workshop: map available data, decision points, uncertainty, and paths to measurable ML value. 3. Vision memo: convert the best ideas into concrete product or model proposals with risk and upside. 4. Scoped proposal: define features, data requirements, timeline, cost, deliverables, and validation criteria. 5. Build and handoff: implement, evaluate, deploy, or transfer the workflow with clear operating guidance. ## Products ### Olva Olva is an invisible real-time meeting assistant for professionals, developed by Mindle and built from Mindle's applied AI product discipline. Product URL: https://olva.ai Olva is a private, desktop-first meeting companion that brings live AI into professional conversations without adding a meeting bot to the call. It works beside Zoom, Google Meet, Microsoft Teams, Slack huddles, Webex, and in-person meetings. It captures desktop audio, detects questions, brings live answers into the conversation, and keeps post-meeting memory searchable. Olva product positioning: - Invisible meeting assistant for professionals. - Desktop-first AI companion for live answers and post-meeting memory. - Designed for professional conversations where privacy, speed, and usefulness matter. - Built by Mindle as the first product featured in the Mindle portfolio. Capabilities: - Live transcription and useful cues - Automatic question detection - AI Boost and document-aware answers - Post-meeting recaps and searchable memory Compatibility: - Zoom - Google Meet - Microsoft Teams - Slack huddles - Webex - In-person meetings ## Proof And Awards Mindle's public proof comes from competitions where every error is measured. The award record includes public leaderboards, prize-winning competition results, patents, and press coverage. Key proof metrics: - $1M Zillow Prize winner: built a state-of-the-art automated home valuation system that beat 3,800+ teams. - Kaggle Grandmaster: global top-tier competition record across ranking, forecasting, valuation, and search. - #1 Canada ranking: ranked #1 in Canada and top 20 worldwide with 10 gold and 10 silver medals. - Zestimate patent: first inventor on a Zillow patent tied to core Zestimate modeling. Source: https://scholar.google.com/citations?view_op=view_citation&hl=en&user=tHeilS8AAAAJ&cstart=20&pagesize=80&citation_for_view=tHeilS8AAAAJ:3fE2CSJIrl8C Award highlights: - $1M Zillow Prize: 1st place among 3,800+ teams from 91 countries. Source: https://zillow.mediaroom.com/2019-01-30-Zillow-Awards-1-Million-to-Team-that-Built-a-Better-Zestimate - Two Sigma Financial Modeling: 2nd place among 2,000+ teams. Source: https://www.kaggle.com/competitions/two-sigma-financial-modeling/leaderboard - Optiver quantitative modeling: top 10 among 4,000+ competitors. Source: https://www.kaggle.com/competitions/optiver-trading-at-the-close/leaderboard - WSDM / KKBox recommendation: 2nd place among 1,000+ teams. Source: https://www.kaggle.com/competitions/kkbox-music-recommendation-challenge/leaderboard - Home Depot search relevance: 2nd place among 2,000+ teams. Source: https://www.kaggle.com/competitions/home-depot-product-search-relevance/leaderboard - Rossmann sales forecasting: 2nd place among 3,738 data scientists. Source: https://www.kaggle.com/competitions/rossmann-store-sales/leaderboard ## Quantitative Finance Proof Mindle's finance work is grounded in public, competitive evidence from Two Sigma, Optiver, and open-source quant ML work referenced by hedge fund practitioners. - Two Sigma quantitative modeling: prize-winning work in market prediction under constrained compute, focused on signal discovery and robust feature evaluation. Source: https://www.kaggle.com/competitions/two-sigma-financial-modeling/leaderboard - Optiver market prediction: top-10 placement among 4,000+ competitors, with open-source quant ML work referenced by hedge fund practitioners. Source: https://www.kaggle.com/competitions/optiver-trading-at-the-close/leaderboard - Research workflow discipline: competition-grade validation habits including leakage control, feature audits, fast experimentation, and benchmark-driven model improvement. Source: https://www.linkedin.com/posts/nimashahbazi_github-nimashahbazioptiver-trading-close-share-7190781939979534337-qzYP/ ## Selected Achievements - Real estate valuation: winner of the $1M Zillow Prize, automated home valuation model, Zestimate patent holder. - Quantitative finance: Two Sigma prize winner, Optiver top-10 placement, open-source quant ML adoption. - Search and recommendation: WSDM / KKBox recommendation challenge, Home Depot search relevance, ranking and relevance modeling. - Forecasting and pricing: Rossmann sales forecasting, Mercari price suggestion, demand and price prediction. - Trustworthy AI: $1M Leaders Prize finalist, AI systems for misinformation, public speaking on trustworthy AI. ## Public Speaking Nima Shahbazi is a public speaker at top-tier AI conferences and applied AI venues. Speaker highlights: - KDD / Canada AI Day: speaker listing and York University coverage for the Canada AI Day trust panel. Source: https://sites.google.com/view/canada-ai-day/speakers - AIC Ontario: Evolving with AI 2026 speaker page and public keynote announcement. Source: https://site.pheedloop.com/event/EVOLVINGWITHVALUE2026/speakers - Georgian: TLIR reinforcement learning applied learning session. Source: https://lu.ma/tlirl-reinforcementlearning - ReWork: Machine Learning Summit Montreal presentation on reducing demand forecast error with deep learning. Source: https://www.re-work.co/events/deep-learning-summit-montreal-canada-track2-2017/speakers - RBC Disruptors: RBC coverage of AI entrepreneurs speaking at its Disruptors forum. Source: https://www.rbcis.com/en/insights/2017/07/rbcdisruptors ## Press And Public Work - GeekWire, Jan 30, 2019: Meet the Zillow Prize winners who get $1M and bragging rights. Source: https://www.geekwire.com/2019/meet-zillow-prize-winners-get-1m-bragging-rights-zestimate-beating-algorithm/ - VentureBeat, Jan 30, 2019: Zillow awards $1 million to team that reduced home valuation algorithm error. Source: https://venturebeat.com/2019/01/30/zillow-awards-1-million-to-team-that-reduced-home-valuation-algorithm-error-to-below-4/ - NVIDIA Developer Blog, Jan 31, 2019: NVIDIA GPUs help developers score $1 million prize for improving Zillow's Zestimate. Source: https://news.developer.nvidia.com/nvidia-gpus-help-developers-score-1-million-prize-for-improving-zillows-zestimate/ - Kaggle Blog, May 25, 2017: Two Sigma Financial Modeling Challenge winner interview. Source: http://blog.kaggle.com/2017/05/25/two-sigma-financial-modeling-challenge-winners-interview-2nd-place-nima-shahbazi-chahhou-mohamed/ ## Contact For consulting, product partnerships, hedge fund ML, applied AI work, products, and services, contact Mindle directly. - Email: sales@mindle.ai - Contact page: https://www.mindle.ai/contact