What we have done is invent the future of BI. Nothing less. Customise the data models. Write the mappings. The Business Requirements Document? This becomes a summary of: Business Objectives.
Establishing credibility? We will demo the applications and show the prospective client what is underneath them. That is, we are not scanning the web for a Business Requirements Template as an example. How are we doing? Best Regards Peter. You know, 16 years ago it was really hard to build a decent DW….
It was really, really hard work. Why are Bills employers not willing to pay for good advice? I am also a member of Inner Circle on DSxchange. It seems a little crazy…… :opinion off. Nick Codling Posted August 17, 0 Comments. Hi Peter, Well said. All the best Nick. This is a template I use to recruit DW candidates for Travelers Insurance in Hartford, CT and St Paul, MN: Because we have several positions…it would help us immensely to know where your strength and interest is…see skills list below If interested in this position, please write back with your availability to interview and to start, the reason you are leaving your present situation OR the reason for leaving your last position, along with your minimum permanent salary requirement and the status of your eligibility to work in the United States?
Also, write back with the number of years of experience and a brief explanation of that specific experience in each of the following: Cognos Cognos 8 business intelligence tools: Framework Manager Report Studio Query Studio Analysis Studio Metric Studio Event Studio [which ones? Ab Initio. Business Objects. Business Intelligence,. Data Mapping. Data Marts. Data Modeling. Data Warehousing. DB2 Documentation [what types of things have you written?
Enterprise Data Repositories. Metadata repositories. SQL queries. SQL Server. Star Schemas. TDWI Teradata experience. Transformation concepts ETL. NET Web Services [which ones? Removed Removed Posted August 22, 0 Comments. Glyn Lewis Posted August 22, 0 Comments. Why am I ranting. Anonymous Posted August 23, 0 Comments. I empathise. Please subscribe. Peter Nolan Posted August 23, 0 Comments. Hi Glyn, Your predicament is common. Been there, done that, got the T-Shirt and Stubby Holder.
Building a DW is so relatively easy now compared to years ago….. Yes, I like the opportunity to prove themselves. Removed Removed Posted August 23, 0 Comments. Glyn Lewis Posted August 26, 0 Comments. Anonymous Posted August 26, 0 Comments. Glyn Lewis Posted August 27, 0 Comments. Hi Nic, Probably not enough info in my last reply. Peter Nolan Posted August 27, 0 Comments.
At this second telco there were 43 financial reports to be migrated. We build a staging area and moved sample data across. And today, we can go even faster than this.
Peter Nolan Posted August 28, 0 Comments. Hi Glyn, So….. And I also wrote…. Anonymous Posted September 4, 0 Comments. It is not the template you are missing, rather an OLAP architect. Register or Login.
Welcome back! Sign in with Email. Reset Your Password We'll send an email with a link to reset your password. Stay ahead! Get the latest news, expert insights and market research, tailored to your interests. Sign up with Email. Sign in with email Enter the email address associated with your account. You auth link is expired or incorrect, please try again. Sign up with email Get the latest news, expert insights and market research, tailored to your interests.
Prospect is usually treated as a status of the customer as a whole, or of the customer for particular products, services or campaigns. Organisation: An organisation is a grouping of one or more individuals, or of one or more companies, or other business activities, or for some other reason are required to be known about.
An organisation may be a legal entity or just some recognised group. Examples of organisations include customers and prospects of Telco products and services, network customers, third-party dealers, the customers of third-party dealers, OLO Other Licensed Operators , competitors, contractors, suppliers, government bodies, Telco own business units, etc.
Individual: An individual is a person for which information can be obtained. An individual may be associated with an organisation, for example as a contact for an organisation e. An individual may also be a customer or prospect for Telco products and services or the customer of a third-party dealer. Customer Churn: Customer Churn is defined as the number of customers leaving Telco from a given level of the customer hierarchy as a proportion of the total customer base at that level presented either as a number or a percentage and calculated at monthly, quarterly and yearly intervals - churn is also calculated by the value of the customer leaving Telco as a proportion of overall revenue.
Dealer: A dealer is a third party organisation that has an existing business base through which it can sell standard Telco Branded Products. Telco owns the customer and consequently the Customer benefits from 24hr customer care and management reports etc. In return for acquiring customers on Telcos behalf the dealer receives an ongoing 'revenue share this is a proportion the customers billing.
Reseller: The Reseller purchases Telco products at wholesale rates and 'resells' them under their own brand. The contractual relationship is between the reseller and the customer, not Telco. The reseller sets his own retail pricing, bills the customer, arranges premise routing and provides customer care. Telco will provide the reseller with a monthly wholesale bill and provide customer care for bill queries and network faults to the reseller only, contact is not made between Telco and the reseller's customer.
The reseller can be further subdivided into two groups: those who aggregate carrier. Aggregator: By using 'smart' MLU technology e. Aggregators are able to maximise their buying power to secure the best possible rate for each route.
Aggregators approach customers as an independent telecomm provider. The offer is to remove the purchase and pricing decision from the customer by pooling rates from all carriers and therefore always offering the best possible combination of rates. Effectively they will manage a customers telephony needs.
Description of limitations of current data warehouse The solution should not support lead tracking, contact management - Not all the information that the sales force will require is likely to be held in the warehouse as this will not be a suitable repository. Prospect and Customer hierarchies - will need to be given as manual or automated input details of group customer s and holding companies. Sources of actual cost information to be investigated and reported on when available.
The Party dimension represents individuals, their parent organisations and the higher-level organisation structure. Additional Descriptive Information This dimension is intended to represent individuals and organisations of interest to Telco so, for instance, this may represent organisations with which Telco have a customer, supplier or interconnect relationship.
The diagram above depicts one of the specific hierarchies that is required the customer reporting hierarchy. Other requirements for the representation of organisation hierarchies may include : Business and trading names of an organisation and all its subsidiaries and parent companies.
Telco defined hierarchies or groupings, such as: o o o o o Sales channel related hierarchies, i. TAL method of grouping customers for a sales account manager. Customer hierarchies defined on an ad-hoc basis for Telco internal reporting. Organisations also include third-party dealers, the customers of third-party dealers, OFTEL, contractors and suppliers.
No hierarchical structure within any of these is required. Data Steward Status Changes Slowly Changing Dimensions The customer reporting hierarchy is likely to change at a rate between quarterly and annually. The Telco sales organisation is likely to change at a rate between monthly for small changes and annually for large changes. Currently available in the Data Warehouse Limitations of current availability Yes. Data Refresh Requirements It is important that this information stays as up to date as possible for those organisations with which Telco carries out transactions.
For other organisations, such as prospects, the data refresh requirement is more lenient. Implementation Notes Customers are the holders of active or deactivated billing system accounts. Additional Commentary As such, this Vanilla Rated Call Revenue Amount is the basic revenue value attached to the call record, and includes factors such as the time of day the call was made, the day of the week, the message source code, the charge band and the specific rating group.
It excludes any subsequent rating changes that are applied by the billing system, which turn Vanilla Rated Call Revenue into the Re-Rated Call Revenue. The Vanilla Rated Call Revenue Amount could also loosely be referred to as unbilled usage revenue, and should therefore have a revenue status type of unbilled usage revenue.
For unbilled usage: Vanilla Rated Call Revenue Amount will be accumulated alongside the Call Duration on a daily basis by telephone number per bill cycle, and will be zeroed at the start of each bill production run for that telephone number. Description of Known Data Quality Problems. Data Retention Requirements Individual Call Detail Records are required for a rolling three months, which therefore implies that Vanilla Rated Call Revenue Amounts will be available for a rolling three months.
The cumulative unbilled usage Vanilla Rated Call Revenue Amount and Call Duration by telephone number by bill cycle will be a daily feed into the system. This should be replaced, in a later increment, by a daily feed of individual unbilled usage CDR records. At the billed usage level, this information should be loaded into the system once per bill production run. That is, there are currently estimated to be six bill production runs per month.
Implementation Notes. The speed of response will depend on the available level of pre-calculated aggregates. This information could be made available from the source systems should the Systems Roadmap point to this being an area of high priority. Of course, every data mart data silo can in theory be sized independently, but ideally, we want all the users to have access to all the data. Why extract, transform and copy the data to another location? Every department should be able to transparently access all the data across the entire enterprise subject to authority , but each independently sized to the task at hand.
Snowflake Computing founded in was established by a team of ex-Oracle database experts, and they have achieved just that. In my next article, I will explain in detail how Snowflake delivers a unique and innovative architecture which supports multiple different sized workloads without any contention.
Each may be sized to the particular problem, and even automatically scale out as additional users make additional demands on the system. A fully ACID transactionally consistent SQL database, it is remarkably easy to maintain with few options to tune and configure — it just works.
Best of all, it is the only data warehouse which was built entirely for the cloud, and has all the benefits of infinite storage, and almost unlimited compute resources on demand. Watch this space! Disclaimer: The opinions expressed in my articles are my own and will not necessarily reflect those of my employer past or present or indeed any client I have worked with. Note: This article first appeared on my personal blog. See the original article here. Thanks for visiting DZone today,.
Edit Profile. Sign Out View Profile. Over 2 million developers have joined DZone. Top 20 Requirements of a Data Warehouse. We take a look at some of the considerations that a great data warehouse needs to cover. Do you agree with these points? Like 2. Join the DZone community and get the full member experience.
Join For Free. The Requirements Workload Separation: One of the greatest challenges facing solution architects today is maintaining the balance of compute resources for a number of competing user groups. Who should be given priority? The diagram below illustrates the massively different workloads of these two competing groups.
This requirement is to separate these workloads, and eliminate the contention between user groups. Maximize Data Loading Throughput: As indicated above, we need to rapidly extract, load, and transform the data, and this means we need to maximize the throughput — the total amount of work completed, rather than the performance of any single query.
Maximize Concurrency: A typical analytics platform has a number of busy users who just want to get their job done. They want their results as quickly as possible, but they are often fighting for machine resources with everyone else.
In summary, we need to maximize concurrency. The ability to handle a large number of queries from multiple users at the same time. Almost every data warehouse, both on-premise and cloud-based, is built upon a single principle: size for the biggest workload, and hope for the best. While solutions like Google BigQuery and Amazon Redshift provide some flexibility, they, like every on-premise platform, are eventually restricted by the size of the platform.
The reality for most users is query performance is often poor, and at month or year end, it becomes a struggle to deliver results in time. In an ideal world, the data warehouse would automatically scale-out to add additional compute resources on-the-fly as they were needed.
0コメント