You have access to more data than ever before -- from your own services and devices, as well as from outside sources -- and you know that there is vast value to be unlocked. But expecting existing IT staff to add this to their existing projects never seems sufficient: you are building new services that don’t look like business-as-usual. FP Complete brings the deep software expertise to deliver the analysis power you know should be possible.
Our software-based and cloud-based systems connect online services, networked devices, unique data flows. We build scalable, reliable, secure online capabilities, with a special emphasis on adding modular capabilities to extract major new value from data.
Whether you are analyzing user behavior, interchanging data between mobile devices, merging and cleansing data, or detecting new patterns in complex dynamical systems such as personalized health or trading networks, you already know that sophisticated software is the right tool. When off-the-shelf applications won’t do the job -- when you have envisioned data-driven services moving your organization far ahead of the way things have been done before, and opening up new features and even new lines of business -- FP Complete has the engineering capability to make those services real.
We can expand upon the capabilities of your existing IT operations and data, to ensure that these new services mesh properly with your existing services. Yet we can also bring to the table, if desired, cutting-edge engineering methods that may be faster and more reliable than traditional projects, and that deliver new levels of maintainability, time-to-market, and scalability.
Social networking giant Facebook needs to keep spam and malware off of its world-leading website across numerous full-sized data centers. A full-time team of programmers is dedicated to this problem, and they require software tools that deliver mission-critical reliability, billion-user scalability, and extraordinary same-day productivity. “Latency is the most important thing. We want to be able to stop attacks immediately,” says engineering manager Louis Brandy. To meet these extreme demands, in 2015 Facebook’s engineering team replaced their anti-spam system with a new one built in Haskell, the core technology of FP Complete’s software. Using the open-source GHC compiler, the team migrated from older tools to Stackage, a Haskell-based tool suite created at FP Complete and maintained by a consortium of FP Complete and the open-source community. The new system automatically uses multi-core parallel processing across a cluster of servers to keep up with Facebook’s ever-growing torrents of data, and new pattern-detection rules are routinely built and deployed by Facebook’s team on a same-day basis. The system can scale up without limit, yet consumes less CPU power per million messages than the old proprietary system it replaced.
A Fortune 500 IT company seeking to better serve one of their own largest customers chose FP Complete to build and deliver a distributed, scalable software-development platform atop their world-class cloud computing service. Starting with existing software developed in-house, FP Complete ported its in-house cloud-based development environment, application deployment system, and software library/package management system, to use the client’s existing application programming interfaces. FP Complete then enhanced the software to take advantage of the client’s existing systems for authentication, fault tolerance (redundant server pools), and scalability (automatic cloud server creation on demand). The complete working system was delivered on time in much less than one year, and immediately validated by the client’s customer.
To serve the open-source community with new developer training capabilities, FP Complete developed, delivered, and operates the School of Haskell online service. This highly scalable Web-based application, built atop the Amazon AWS cloud, has been running for two years. It regularly serves over 1000 users per day with online programming lessons, including the ability to actually edit and run programs in secure containers on its own shared Ubuntu Linux servers. Currently hosting about 500 separate lessons, the system automatically adapts to bursts in demand when major new tutorials are published, detecting high workloads and spinning up additional replicated servers with no human intervention. The system is so reliable that even with its demanding and experiment-prone user base, zero dedicated staff are required to keep it in continuous 24x7 operation.
To arrange a complimentary conversation around your data and analytics needs, and to determine which capabilities we can help with immediately, email us at firstname.lastname@example.org. We look forward to hearing from you.
Our Team and Qualifications
Our team have built and delivered solutions in areas including:
- safety-critical FDA-regulated pharma
- secure elections (extremely wide-area voting and government-regulated tabulation)
- aerospace vehicle design
- forensic data management
- credit card processing
- ultra-high-volume online services
- ultra-high-volume packaged software
Included in our team are:
- a former head of software tools for Microsoft
- several PhD computer scientists with expertise in machine learning, mathematical modeling, high-performance computing, and mixed-language programming
- experts on dynamical systems
- experts on high-performance computing (HPC) scalable system operations
- experts in cloud quality assurance and security
- engineers from some of the largest online data management and analysis systems at Microsoft and Google.
As an onshore provider with highly qualified staff and project leaders, we are able to take responsibility for work in highly controlled project settings, from secure proprietary systems to government-regulated and published products.