AI that learns your MultiValue system and answers questions about it
Trawler explores your UniVerse, UniData, jBASE or D3 applications, writes the documentation nobody has had time to write, and answers your team’s questions in plain English. It runs inside your network.
- Documents every program, file and dependency, and keeps it current as your code changes
- Answers questions from your real code, and says so when it doesn’t know
- Stays on your servers: read-only, nothing installed on your systems
Decades of business logic, understood by a few people
Your MultiValue applications keep the business running. Over time developers move on, documentation falls behind, and the reasons behind the code get harder to find.
“After thirty years, I’m retiring. My last day is two weeks from today.”
The email every MultiValue shop eventually getsKnowledge lives with a few people
Trawler captures how your applications work, so it doesn’t leave when your experienced developers do.
Documentation is missing or out of date
It documents programs, files, dependencies and business rules, and updates them as your system changes.
Troubleshooting takes too long
Ask about your applications and get answers grounded in your actual code and environment.
Changes carry hidden risk
See which programs, files and jobs a change touches before anyone edits production code.
New developers face a steep climb
New people can explore the system on their own instead of waiting for the one person who knows.
Ask it what you’d ask your best developer
Real screens from Trawler, working on RFMS, MultiValue Central’s demo rental system on UniVerse.
From one program to your whole environment
Point Trawler at as much or as little as you want explained.
A program
What it does, which files it reads and writes, and the business rules inside it.
An account
How its programs, files, dictionaries and VOC entries fit together.
A server
Accounts, scheduled jobs, phantoms and the services around them.
Your environment
The servers, applications and dependencies that support the business.
CALL RF.SESSION.INIT
CHAIN ‘RF.BILL.RUN’
How it works
Two ways to put it to work: documentation that stays current, and answers whenever someone needs one.
Connect
Trawler gets read-only access to the systems you approve. Nothing is installed on them.
Learn
It explores programs, files, schedules and dependencies and writes down what it finds. For the first month our MultiValue engineers work alongside it, pointing it at what matters most to you.
Ask
Anyone on your team can ask in plain English. When it doesn’t have enough to answer, it says what’s missing instead of guessing.
Everything it learns, from 30 years of code to the tribal knowledge nobody wrote down, becomes your knowledge base: documentation your team can ask questions of.
AI-powered understanding. Your network. Your control.
Your application knowledge stays on infrastructure you control, instead of being sent to an outside AI service.
Read-only by default
Discovery uses pre-approved operations and installs nothing on the systems it examines.
Secrets removed before any AI sees them
A fixed redaction step, not an AI, strips passwords and keys from everything it collects.
Runs on your servers
We size the servers with you and set them up as part of the engagement.
Built for the MultiValue systems you rely on
General-purpose AI tools know very little about MultiValue. Trawler is built for it, and it learns your system, not a generic one.
Questions IT asks first
Does our code or data leave our network?
No. Trawler runs on servers inside your network, and what it learns is stored there. It contacts MultiValue Central only for licensing and software updates.
Will it change anything on our systems?
Discovery is read-only by default. It uses a fixed set of pre-approved operations and installs no software on the systems it examines.
What happens when it doesn’t know the answer?
It says so, and names what it would need to find out, instead of guessing. Answers come from evidence it read on your systems.
Which MultiValue platforms does it support?
Rocket UniVerse, UniData, jBASE and D3, along with the programs, dictionaries, PROCs, phantoms and jobs that run on them.
What do we need to run it?
Trawler runs on GPU servers inside your network. We size them with you and set them up as part of the engagement.
How long before it’s useful?
Most of the learning happens in the first month, with our MultiValue engineers working alongside it on the parts of your system that matter most to you.
Capture what your system knows while the people who know it are still here
See Trawler on a MultiValue system like yours. A 30-minute call with our team, no slides.