Urvesh Devani Email & Phone Number

Singapore

Building AI-driven systems that empower shippers to optimize costs, enhance customer service, and streamline invoicing and auditing.Expertise in freight audit automation using AI-based OCR, reliable prediction engine for container milestones, and developing scalable analytics infrastructure to support operations and mitigate risks

Singapore

- Built machine learning and analytics backbone for real-time predictive supply chain visibility. Automating freight invoice processing using LMM - Collaborated actively with customers and internal teams to identify and quantify data quality issues with supply chain data. Developed new KPIs to address and monitor these issues and automated QA - Developed ETA (Estimated Time of Arrival) Prediction Engine, Forecasting tools, Interpretable ML features, Anomaly detection and ranking engine - Streamlined analytics and BI processes with an end-to-end pipeline and SOPs, slashing data query response times by over 90% - Expertise: Hands-on experience in Generative AI applications, interpretable machine learning, time series forecasting, cloud services (AWS, Google Cloud), ETL, Data Warehouse, BI, with proficiency in Python and SQL- Helped develop product strategy, incorporating feedback from key industry players like shippers, freight forwarders, and carriers to inform feature development of innovative solutions including sustainable supply chain, risk management and recommendations.- Directed a diverse, remote team of data analysts, engineers, and scientists, fostering a culture of innovation and continuous improvement

Singapore

- Started working with founders from early stage of the company (seed round-team of 4 full time including founders)- Helped company in exploring product market fit, understanding customer pain points, shaping-scaling product and growing data science/tech team- Built Demand forecasting Engine for cargo flows for ocean as well as air cargo. Worked on explainable ML module to show impact of macroeconomic indicators. Worked closely with engineering team to make KPIs intuitive and explain statistics on UI that hosts thousands of time-series forecasts.

United States

spent a few hours every week mentoring students of Data Science Career Track. Assisted 10 people achieve their goals in data science learning.

- Delivered value to enterprise clients by building prediction engines(Steel manufacturing, Private jet airlines)- Helped the pre-sales team to pitch ideas, in preparing demos- Prepared training materials- Contributed towards AI product development- Mentored interns, new hires

Worked as a part of the team to help a steel manufacturer optimize their schedule and deliveryhttps://www.forbes.com/sites/tomdavenport/2018/02/08/shining-up-a-rusty-industry-with-artificial-intelligence/

Bangalore

- Have been part of an early-stage team that delivered for the first client of noodle.ai. The team dealt with social network big data (in range of TBs). Built, evaluated, and shipped recommendation system.Unsupervised learning approaches, Content and network graph-based models, Geolocation Prediction, Experiments: Human Evaluation of recommendations- Developed a composite index and a risk engine as hedging tools to address the problem of raw material price volatility.- Developed a demand prediction engine for a Private Jet airline.https://flyxo.com/blog/xojet-noodle-ai-shaping-future-dynamic-pricing/

Bengaluru Area, India

- Wrote code(c,c++) for various modules of Next Generation FireWall including NAT, DHCP, Syslog - Design, maintenance of modules, Unit and Integration testing - Addressed high severity customer queries and issues- Collaborated with network security research teams to implement machine learning models for Malware classification, Intrusion Detection, Prevention System https://github.com/stratosphereips/StratosphereLibSlips- Part of ThingQbator (Internal IoT makerspace), Prepared Internal surveys that help executives make decisions

Mumbai Area, India

Master Of Technology (Mtech), Software Engineering, 9.5

B.E., C.S., 8.35