Santiago is available for hire
Hire SantiagoSantiago Hernández
Verified Expert in Engineering
机器学习工程师和开发人员
Location
布宜诺斯艾利斯，阿根廷
Toptal Member Since
January 29, 2019
Santiago是一名热衷于解决大规模数据问题的机器学习工程师, optimization, 工程问题. He has led a team of data scientists to build products that predict the value of over 300 million users worldwide to optimize the advertisement of world-leading companies. 圣地亚哥认为发展优雅是一项严肃的事业, efficient, 以及精心设计的软件系统.
Portfolio
ZeeMaps (via Toptal)
Bootstrap, PostgreSQL, Redis, Plotly, ipythnotebook, Docker, SaaS, Dash...
自由市场(包含Mutt数据)
Docker，机器学习，ippython笔记本，优化，编程...
Jampp
Pandas, Scikit-learn, Tornado, C, Presto, PostgreSQL, Bash, Git...
Experience
Availability
Part-time
首选的环境
Bash, Vim文本编辑器，Python, Git, Linux
The most amazing...
...machine learning system I've developed streams over 60 million messages per hour and builds datasets in real time to predict monetization on mobile apps.
Work Experience
Data Scientist
2019 - PRESENT
ZeeMaps (via Toptal)
- Analyzed the company's system-as-a-service (SaaS) data to find insights and make strategic business decisions.
- Built a dashboard to evaluate the company's evolution and easily see the analysis and aid decision making.
- Classified subscribed users into acquired and at-risk users—enabling us to focus on preventing risk users of churning, 推动公司发展.
- Provided visibility on the revenue, lifetime value, churn rate and other metrics per user category.
- 标记高和中等流失风险的用户，以便提供重点支持和客户服务.
- 分析和优化赞助搜索活动，以获得新用户.
- Dockerized仪表盘系统的设置和部署.
- Performed cost analysis per plan and per user to evaluate the current SaaS's pricing methodology and improve the revenue by performing deals with high consumers when appropriate.
技术:引导, PostgreSQL, Redis, Plotly, ipythnotebook, Docker, SaaS, Dash, Python, Data Science
机器学习和优化领导者
2019 - 2020
自由市场(包含Mutt数据)
- Defined an optimization problem to optimize the marketing's team budget and target ROAS (return on advertising spend) goal allocations revenue from Google Shopping.
- Performed an exploratory data analysis on the company's and the marketing's team data in order to understand the relationship between budgets, goals, and results.
- 开发了成本预测和预测系统, revenue and return of investment of Google Shopping campaigns depending on the budget and target ROAS goal subject to business constraints.
- Engineered and developed a system to solve the optimization problem and define how to set up digital marketing campaigns on Google Shopping based on the predictions, 预测和业务限制.
- 使用docker容器和docker compose对系统的设置和部署进行Dockerized.
- 在一个国家为两个账户部署了系统, after the results, 在同一个国家的另外五个账户上, 随后推广到其他国家.
Technologies: Docker，机器学习，ippython笔记本，优化，编程, Prophet ERP, Pandas, Scikit-learn, Forecasting, Data Science, Python
机器学习工程负责人
2015 - 2019
Jampp
- Built machine learning online estimators processing over 60 million programmatic advertisement messages per hour (an auction's win rate, 第二种价格是拍卖的成本, 以及新用户和应用内事件的转换).
- Performed optimization and feature engineering on models leading to over 15% in conversions and a 30% increase in the company's net revenue.
- Developed a revenue and inventory purchase optimization system that resulted in over 20% increase in the net revenue.
- Created several web interfaces to provide visibility and interpretability to machine learning and optimization systems.
- Led data scientists to define and develop user clustering systems processing over 200 million users.
- Developed a system that uses the previously mentioned ones to make various decisions—like how much to bid and for which client—during real-time auctions selling advertisement slots on mobile apps.
Technologies: Pandas, Scikit-learn, Tornado, C, Presto, PostgreSQL, Bash, Git, 亚马逊网络服务(AWS), Linux, Python
研究员|开发人员
2013 - 2014
综合神经科学实验室
- Created a mind speller in Python using Fisher’s LDA and SVMs compatible with the EMOTIV EPOC electroencephalography headsets.
- 在一个基于放松的竞赛游戏中工作，使用阿尔法脑波检测玩家.
- Contributed to the development of a steady-state visually evoked potential selection interface to control Lego Mindstorm cars from afar.
- 建了一个网络服务器来控制耳机, 启动不同的系统, 记录大脑活动和用户输入来扩大我们的数据集.
- Researched on mind-speller variants using different classification algorithms and signaling environments.
技术:Emotiv SDK, Python
Experience
建立更强大的数据科学团队
Nowadays, 在数据科学团队中工作需要与来自不同背景的人一起工作, 比如数学家, economists, actuaries, physicists, and engineers. And being a technical leader of a data science team means that not only you have to ensure that the research, insights, 产品不仅为公司增加了价值，而且还具有可复制性, maintainable, reliable, scalable, performant, testable, and correct.
For this reason, I started a series of on-site technical pieces of training in computer science and software engineering because the team's data scientist did not have an engineering or computer science background.
培训课程的主题是算法复杂性, programming exercises, environment setups, data structures, 面向对象的哲学, 软件架构.
On the tech side, this—in conjunction with proper onboardings and code reviews—led to better products and faster development times. On the social side, we were happy to find out that we were unexpectedly making a stronger team by the bonding that arose from studying and debating shared interests.
For this reason, I started a series of on-site technical pieces of training in computer science and software engineering because the team's data scientist did not have an engineering or computer science background.
培训课程的主题是算法复杂性, programming exercises, environment setups, data structures, 面向对象的哲学, 软件架构.
On the tech side, this—in conjunction with proper onboardings and code reviews—led to better products and faster development times. On the social side, we were happy to find out that we were unexpectedly making a stronger team by the bonding that arose from studying and debating shared interests.
数字营销的动态消费模型
在数字营销行业, performance marketing platforms and their customers have to define the advertisement campaign's goals and a spend model to determine how the customer will be charged (like a fixed CPM, CPC, CPI or CPA).
I worked on a new spend model in which the client and the platform agree on a maximum budget and goals—such as a maximum cost per click or purchase—and the system finds the optimum price to charge the client while achieving the goals and ensuring satisfactory results for both parties.
I worked on a new spend model in which the client and the platform agree on a maximum budget and goals—such as a maximum cost per click or purchase—and the system finds the optimum price to charge the client while achieving the goals and ensuring satisfactory results for both parties.
移动应用程序中的事件预测
I built a statistical learning system to estimate the probabilities of mobile app users performing events on applications. 在线算法每天处理超过400万个数据点，以不断拟合新数据.
This was a challenging project as there were different types of apps (like food delivery or flight reservation apps), 不同类型的事件(如搜索或购买), and delayed conversion feedback (people usually do not book a flight right after watching an ad).
在将其部署到生产环境并将其用于收益优化之前, 回答以下几个问题是很重要的:
•适合一个好的估计器所需的最小数据量是多少?
•如果你停止从数据中学习，继续使用当前的估算器，会发生什么?
•你有多愿意等待延迟的反馈?
•在应用程序中的所有事件中，你想预测哪一个?
在解决了问题并考虑了可能的突发事件之后, 该系统顺利推出，没有出现任何问题.
This was a challenging project as there were different types of apps (like food delivery or flight reservation apps), 不同类型的事件(如搜索或购买), and delayed conversion feedback (people usually do not book a flight right after watching an ad).
在将其部署到生产环境并将其用于收益优化之前, 回答以下几个问题是很重要的:
•适合一个好的估计器所需的最小数据量是多少?
•如果你停止从数据中学习，继续使用当前的估算器，会发生什么?
•你有多愿意等待延迟的反馈?
•在应用程序中的所有事件中，你想预测哪一个?
在解决了问题并考虑了可能的突发事件之后, 该系统顺利推出，没有出现任何问题.
理财教育手游
I designed and built the back-end of a mobile game oriented to provide financial education to teenagers.
这款游戏从头开始模拟说唱歌手的职业生涯, in which the player had to give concerts playing mini-games and take good decisions in order to progress in their career.
球员的经理通过给他建议和建议来平衡业余时间来帮助他, 借钱去更大的场地玩游戏，购买物品来增加角色的流量. 该项目在Itau银行和Red Hat组织的竞赛中提出.
这款游戏从头开始模拟说唱歌手的职业生涯, in which the player had to give concerts playing mini-games and take good decisions in order to progress in their career.
球员的经理通过给他建议和建议来平衡业余时间来帮助他, 借钱去更大的场地玩游戏，购买物品来增加角色的流量. 该项目在Itau银行和Red Hat组织的竞赛中提出.
Education
2015 - 2019
计算机科学硕士学位
布宜诺斯艾利斯大学-布宜诺斯艾利斯，阿根廷
2009 - 2014
计算机科学学士学位
布宜诺斯艾利斯大学-布宜诺斯艾利斯，阿根廷
Skills
Libraries/APIs
Scikit-learn, Pandas, REST api, NumPy, SciPy, Matplotlib, ZeroMQ, Keras
Tools
Git, Docker Compose, Vim Text Editor, Emotiv SDK, ippython Notebook, Plotly, Prophet ERP
Languages
Python 3, Python, SQL, Bash Script, Bash, C, HTML, JavaScript
Frameworks
很快，Flask, Bootstrap
Storage
PostgreSQL, Redis, Amazon S3, MySQL
Platforms
Linux、Docker、Jupyter Notebook、亚马逊网络服务(AWS)、Amazon EC2
Paradigms
Data Science
Other
点击率(CTR), Machine Learning, Tornado, Algorithms, Distributed Systems, Optimization, Vowpal Wabbit, SSH, Multiprocessing, Dash, SaaS, Forecasting, Programming, Image Processing
有效的合作
如何使用Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
在与Toptal领域专家的电话中讨论您的需求并细化您的范围.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
开始你的无风险人才试验
与你选择的人才一起工作，试用最多两周. 只有当你决定雇佣他们时才付钱.
对顶尖人才的需求很大.
Start hiring