Google于2005年在Belo Horizonte落户，靠近UFMG（米纳斯吉拉斯州的联邦大学），这是世界上最负盛名的信息检索研究实验室之一。它背后的故事值得一讲：由Berthier Ribeiro-Neto领导的一组研究人员于1999年创建了一个名为“ TodoBr”的搜索引擎。随后，该团队成立了一家名为“ Akwan”的初创公司，以将TodoBr商业化。
原文题目：Why You Should Source Foreign Talent for AI, IoT and ML
原文：The artificial intelligence revolution has been going on for a while, but that does not mean all of software research and development is part of it. It is, nonetheless, an interesting example of what pushes the software industry forward.
The ideas and techniques that brought machine learning into existence were developed over the span of several decades, gaining momentum in the 80s as a long-term endeavor that started in academia. Significant development occurred once the community started comparing its tools in competitions, and measuring performance and applicability.
Fast-forward to the late 90s, and the big software companies would start to develop techniques and infrastructure to gather, store and process huge amounts of data -- data that then could be used to train classification and search-based algorithms.
With the addition of deep neural networks, what was at some point a rudimentary set of academic tools turned into an industrial powerhouse. Targeted advertising, consumer behavior analysis, and many other areas benefited from this ecosystem. Data engineering started turning information into utilities.
The advent of cryptocurrency and blockchain followed a similar path. What started as academic work became an important milestone even well beyond the software industry. What is somehow contradictory is that we are now harvesting the bounty of automated reasoning and complex software systems in general, at the cost of hyperspecialization.
Technology needs supervision, and specialized technologies need specialized and highly skilled resources. There is no such thing as a free lunch. In an ever-changing, competitive market, the availability of human resources is key to the success, if not the proper operation, of your data-driven business.
The convergence of several factors, and the chiseling hand of time, have turned Latin America into a sweet spot when it comes to headhunting. Latin American universities offer a wide range of software related degrees, from undergraduate to doctoral, particularly in Brazil, Argentina and Chile.
The software industry has been growing strong over the decades, and a proper software culture came into fruition once the researchers and engineers who had built their careers across top-level companies and laboratories brought back knowledge, networking and best practices with them, nourishing the local companies and research groups.
Unprecedented development has taken place in the software and engineering ecosystem in recent years. Both the public and the private sectors have been investing in software-related institutions and projects, with a particular focus on automated reasoning, neuroscience and software engineering.
Adding the region's convenient exchange rates to the equation leads to the conclusion that Latin America has become a good place to look for specialized skills.
Taking a closer look at Brazil, we can point to the São Paulo Research Foundation, or FAPESP, which is a public foundation with the mission to support research projects in higher education and research institutions.
It has seen a disbursement in 2017 of R$1.058 billion (approximately US$283 million) with 38 percent of the disbursement going to supporting advancement of knowledge, 5 percent to supporting research infrastructure, and 57 percent to supporting application-driven research -- in many cases performed in small businesses or in a joint effort by academia and industry.
They are now about to launch an AI research center in São Paulo, together with IBM, that will receive $20 million in investment, coming from both industrial and academic organizations that may join the initiative as partners.
Future development in the area is to be expected, not only in public and private cooperation scenarios such as FAPESP-IBM, but also through the efforts of big players such as Bradesco, the second biggest bank in Brazil, which launched Next. A digital bank focused toward millenials, Next expects to improve risk and credit analysis, customer experience, and security by investing in artificial intelligence.
Google settled in Belo Horizonte back in 2005, close to UFMG (Federal University of Minas Gerais), one of the world's most prestigious research labs in information retrieval. The story behind it is worth telling: A group of researchers, led by Berthier Ribeiro-Neto, created a search engine in 1999 called "TodoBr." The team later founded a startup called "Akwan" to commercialize TodoBr.
For years, Akwan survived without external capital. A Google executive got to meet Berthier and liked the way Akwan was built and operated so much that he convinced Eric Schmidt that buying Akwan would be better than building an operation from scratch. Seven months later, Google acquired Akwan and made Berthier engineering director for Latam.
In Argentina, the software community has been growing strong both in the industry and academia. In 2015, the International Joint Conference on Artificial Intelligence (IJCAI) was held there; in 2017, the International Conference on Software Engineering took its turn.
Several key laboratories are located there, such as the Laboratory on Foundations and Tools for Software Engineering (LaFHIS) from the University of Buenos Aires (UBA), where researchers are working on formal methods for software engineering, model checking, static and dynamic analysis, and automated tests generation.
The Applied Artificial Intelligence Laboratory (LIAA), also at UBA, is an interdisciplinary environment, which combines different aspects from computational neuroscience -- such as complexity and randomness perception in humans, computational linguistics, data mining in big text corpus and source code, interactive dialogue systems, speech recognition, and real-time analysis of brain signals.
The researchers developed Eva alongside Workia, a company that specializes in developing products and services based on technology for the management of human resources. Eva is a virtual interviewer who, through machine learning and artificial intelligence, can carry out personnel selection interviews.
The virtual interview allows the team to analyze the discursive coherence of a candidate and seven technical competences, among them proactivity and leadership. It can be used as a first analysis, it is complementary, and it has been reported that in eight out of 10 cases, the algorithm is evaluating competencies as an expert interviewer.
Chile has shown to be particularly fruitful in the development and adoption of the Internet of Things (IoT). The Chilean company Satelnet has been making it possible for its clients to run operations in remote locations in the south of the country, and to access online information even on mobile bases. They can view and control remote cameras or run any production process automatically through satellite technology.
Customers can have immediate access to the information that is being gathered from the equipment found in in Magallanes pontoons, fish farms, or boats, in order to aid decision-making.
Satelnet implemented a solution that benefits from the IoT infrastructure by transferring reliable, well-structured information, regardless of the distance and nature of its source. Several companies have been hopping onto the IoT wagon in Chile, implementing it for their own processes or developing solutions for third parties -- as is the case with Satelnet, Fujitsu and others.
We can expect the region to keep pushing the software industry forward. It has the research, the development and the adoption. The exchange rate makes it profitable for highly skilled resources to initiate business with companies in the U.S. and overseas.
Latin-American resources are highly skilled, affordable, and tend to have a keen eye for problem solving and creativity. Key to the success of these collaboration opportunities is communication and understanding.
Once outsourcing begins, there are several things to take into account. First, you will have to share information about your organization, your corporate culture, your vision, your mission, and what you expect from your team. Plan your meetings -- try to find some time for your local and your remote teams to chat face-to-face on a regular basis.
If you are located in the U.S. and your remote team is in Latam, the time zone should not be an issue. Remember that language can be a barrier at first, that some things may be lost in translation and due to cultural differences, but with a good disposition this will improve greatly over time.
Pay attention to references, idioms, and even body language that could be specific to your region or culture. Some denaturalization may be in order. You need to make sure that you both understand each other and are on the same page when it comes to business.
The not-so-marginal benefit of this is that you may be forced to review your communication protocol. You may want to share documentation not only about the business-specific information, but also about your processes.
If this is your first time with a remote team, it is important to have a mechanism to keep track of your goals, identify any obstacles that may arise, and allow clarification of any particular topic. It is essential to keep a record of what has been discussed and agreed upon on each side, so that you then can share relevant information. Failing to do so can lead to misunderstandings and wasted opportunities.
You may be tempted to think that something was poorly executed when it may have been poorly coordinated or communicated. Keeping an eye on those incidents can help you improve your overall business processes, not only when working with a remote team, but also locally. It will force you to review your vision, your goals, your corporate structure, culture and communication capabilities.
When it comes to the more mundane aspects of collaboration, make sure that you have enough bandwidth on both sides to allow for proper video and audio transmission. If you are being charged or are bookkeeping on a per-hour basis, find the time-tracking system that best suits your needs, and make sure that your remote team knows how to use it.
It is not a big deal to have a slow start, as long as you can refine and improve your processes and understanding of each other. It will pay off in the midterm. Try to have an approach that is as transparent as possible, both with your local and remote teams. It goes without saying that you must be open to questions.
Try to ask if things are understood and if your crew can identify or foresee any obstacle or missed opportunity in the future. Review the requirements, data dictionaries, documentation and even code together, if possible. The game is not only about being open, but also about developing a critical and business-focused frame of thought.
In a nutshell, keep in mind that specialized manpower is key to running a successful business, and communication is key to outsourcing successfully.
Do not miss the opportunity to try and incorporate Latin American resources into your staff. Make them part of the pipeline. Take advantage of the cultural and knowledge exchange, increase your throughput in critical domains, improve your processes, and expand.
State Abstraction as Compression in Apprenticeship Learning https://github.com/d...
? 腾讯ISUX isux.tencent.com 社交用户体验设计 ? ISUX实习生招聘2019专属通道 ? ISUX 2019年实习生招聘正式启动，爱...
I took part in the summer school hosted by Harbin Institute of Technology. The T...