Agentic systems
We build custom multi-agent AI systems on your documents, records and data. Each agent has one job, and an audit step checks every citation against its source before an answer reaches you.
What we build
Systems for evidence-heavy work over large bodies of text and data, shaped around the way your team already works.
Research and analysis agents
A planner breaks a question into parts and gives each part to a specialist agent. Their findings are audited and combined into one answer with sources. Suited to legal research, due diligence and pricing analysis.
Document and regulation extraction
Agents read contracts, filings, rulebooks and regulation, and return structured records. Each field links to the passage it came from.
Retrieval with checked citations
Question answering over your documents. Each citation in an answer must match a passage the system actually retrieved, or the claim is removed.
Internal assistants
Assistants that work inside your systems with each user's own permissions, and keep a log of every action they take.
Governed tool access for AI assistants
MCP servers that connect AI assistants to business tools under rules you set for tools, accounts and actions. Every call is recorded.
Agents over Rillor datasets
Our datasets connect as tools: market data, GPU and neocloud pricing, legal datasets and public procurement records, next to your own data.
How a system fits together
Every Rillor agent system follows the same pattern. Each part has one job, and every step is logged so a run can be replayed.
- Planner
- Reads the question, splits it into parts and assigns each part to a specialist agent.
- Specialist agents
- Each works one part, with its own instructions and only the tools that part needs.
- Tools and data
- Retrieval over your documents, your databases, Rillor datasets and APIs, reached through governed tool access. Every result is logged as evidence.
- Audit step
- Compares every citation in the findings with the logged evidence. A claim without matching evidence goes back for rework or is removed.
- Answer
- Delivered with its citations, each one checked against the source passage.
- Models
- Model families are chosen per project and versions are pinned, so every run can be replayed.
Running in production
Two of our own systems are built on this pattern.
Grellum
Agentic legal research. In deep mode, a planner assigns each question to four to six specialist agents, and an audit step reviews their work. Every citation must match retrieved evidence.
Every citation checked against the evidence it came from
MegaMCPs
Governed access to business tools for AI assistants. Each customer gets an isolated MCP server, and no tool runtime is shared between customers.
One isolated MCP server per customer
Both run on the Rillor platform, where tenant isolation was verified by 109 access, admission and network tests on 7 Oct 2026. See all systems
Tested, then deployed privately
A system goes live once it passes tests built from your own material, in an environment you control.
Evaluation before go-live
We test each system on questions with known answers drawn from your material, plus out-of-scope and adversarial questions. We measure answer quality, citation accuracy and failure cases for each model family and version, and log every run so it can be replayed.
What an evaluation report covers
- The test questions and the answers agreed during scoping
- Citation checks against the source passage
- Failure cases, each with the logged run behind it
- Results by model family and pinned version
Deployment on private AI cloud
Agent systems run in a dedicated environment on your servers, in your colocation space, or on GPU capacity Rillor sources for you. Your documents, prompts and logs stay in that environment.
What runs in the environment
- Model serving for the model families you choose
- Retrieval indexes over your documents
- The agents, their tools and the audit step
- Access control and audit logs
How a build runs
Four stages. You review the result of each one before the next begins.
- ScopeWe map the workflow, the sources and the decisions the system supports, and agree what a correct answer looks like.
- BuildWe build the agents, tools and retrieval over your data, with every step logged.
- EvaluateWe run the agreed tests and failure cases, fix what fails and report the results.
- DeployWe deploy into your environment and operate the system, or hand it to your team with runbooks.
Notices
AI outputs. AI outputs can be wrong. Important decisions should include human review.