Two new technologies that allow everyone to handle data
To achieve our vision of ‘making data an everyday tool’, we are developing a new data management tool.
Morph is based on the concept of an ‘all-in-one Data Studio,’ and makes easy and powerful use of data available to not just engineers but also marketing, sales, customer support, and management.
The challenges we face: data should be easier
We have been building business software and end-user software for various companies for seven years. This included projects that dealt with complex data, such as machine learning and blockchain.
We became aware of the major challenges faced by all business people who work with data, and we are trying to solve them.
1. The need for an engineer
First, and perhaps most obvious: data is very difficult to handle, so you have no choice but to ask a knowledgeable engineer to extract the data for you, or build a dedicated system.
Customers have asked us, “can you output the data in CSV so we can use it?” I would assume that many of our readers may have made or received a similar request. This challenge worsens as the amount of data available to businesses increases.
2. While database technology has evolved, database clients have not evolved at all
The development of database technology for utilizing data has been progressing all over the world and many innovations have been made. Large data and complex aggregations that cannot be processed by Excel or Google spreadsheets can be easily handled using databases.
However, the use of databases is currently only available to engineers, as advanced skills and knowledge are required to use such technology.
In other words, database technology is evolving at a tremendous pace, but client tools are not evolving at all. There are no client tools that can create new databases, connect to existing databases, extract / save data, or create APIs and visualizations in one place.
3. The cost of implementing a database is too high
A database is required if you want to manage a reasonable size of data within your company. Databases are expensive to use, and require advanced knowledge for selection and construction. In most cases database selection and construction is not done by the department using it, but rather by consulting internal and external engineers. As mentioned above, there are no client tools other than those for engineers, so it becomes necessary to develop the interface to use the database as well. This also leads to vendor lock-in.
Thus, the high cost of implementing databases has further raised the hurdles to database utilization.
Why the solution is now possible — two technologies.
1. Serverless RDB (serverless relational database)
In recent years, database technology has seen innovations not only in scalability, but also in the way it is hosted.
We are developing Morph using the open source Serverless Postgres system called Neon, which automatically puts normally expensive database instances on standby when they are not in use. This greatly reduces the cost of database usage. Users can also dynamically adjust start-up when using the database, so the database uses only the amount of resources they need.
Neon also automates all the work involved in setting up a database, so you get a database with one click where you would normally have to log in to a cloud platform such as AWS, Azure or GCP and do a lot of configuration before you can build it.
Neon’s technology is also used in the recently announced Vercel Postgres, which shows the reliability of Neon’s technology in this area.
2. Natural language interfaces (LLM and Chat GPT)
LLMs, including ChatGPT, play an important role in enabling new data tools.
Until now, in order to manipulate databases programming languages such as SQL and Python had to be learnt. This knowledge was a huge hurdle for people who do not know how to code, or how to code in the back end.
We thought that Chat GPT might be the solution, and through repeated experimentation we discovered that it is very good at outputting programming languages.
We have integrated Chat GPT into Morph. Even in the current early version, with natural language instructions we are able to perform:
- Data extraction (automatic SQL generation and execution)
- Visualization and complex data analysis (automatic Python generation, execution)
- Structuring unstructured data to match the table structure in the DB
…as well as several other processes that previously required a high level of technical expertise.
With the emergence of code-specific LLMs such as StarCoder, this functionality can only be expected to develop. For our part, we aim to bring the knowledge required to use the database closer to zero through multimodal support and the development of dedicated LLMs.
What kind of tool is Morph?
Morph is a no-code tool for data management. Without any programming knowledge, it can be used to:
- Create new databases
- Import existing databases (MySQL, PostgresSQL)
- Import CSV, Excel files, etc.
- Display, manage and co-edit records
- Create APIs to manipulate databases (see API Document for details)
- Aggregate, analyze and visualize data.
Request
Data should be easy for everyone to understand and use, and this should be completed through one simple interface.
Morph is a tool for data operations, but the processing and aggregation of data has a wide range of applications and expectations depending on the business.
However, through internal testing — and from prior users in a wide range of industries, including tech companies, manufacturing, call centres and healthcare — it has become clear that these processes are very similar in nature.
To achieve our vision of completing all data work on one simple interface, we need feedback and input from people from a wide range of industries and backgrounds.
We invite you to join us on our journey by following the links below.
Discord: https://discord.com/invite/8ZcSbDrN6e
Try for free
Morph has a free plan which you can start using right now. Please try it out!
