Skip to content
Technology

The technology under Agent Hive. Seven ideas behind the product.

Seven problems we work on. Each figure runs in your browser, and most can switch between the idea and how it appears in the product.

01 · Context factory

Sources in, reviewed context out. Built once, reviewed by people.

Databases, documents and images become one reviewed catalog graph, with lineage between every table and KPI.

A synthetic catalog. The tour opens a table and traces a KPI’s lineage.Illustrative data
Every format, one space

Text, images, tables and audio, side by side. Matching items are pulled together.

Contrastive alignment, simulated in 2-D. Hover an item to see its matches.Illustrative

How a question finds its context

A question lands near its answers. Then the closest few are read.

  1. 01 · EmbedAsk

    “Why did on-time delivery slip in Q3?” becomes a vector, like every table and page in the workspace.

  2. 02 · CompareMeasure

    Similarity to everything in the workspace. Each ring is one level.

  3. 03 · RetrieveRead the closest

    The top five reach the agent, each with its source.

    • Delay root cause · recipe0.93
    • On-time delivery % · KPI0.92
    • delivery_events · table0.89
    • Avg delay (days) · KPI0.85
    • Delivery KPI definition · wiki0.83
02 · Ontology

An ontology with room to grow. Hierarchies live naturally in hyperbolic space.

A hierarchy grows exponentially with depth, and so does a hyperbolic disk, so it fits with room at every level.

The hierarchy spreads toward the rim. Drag to bring any branch to the centre.Illustrative data
03 · Agent swarms

Many small agents, one good path. Coordination through the graph itself.

Agents mark the edges they use, useful marks are reinforced, and a path emerges that no single agent planned.

Agents deposit on useful edges, and the answer path emerges.Illustrative data
04 · Long-running agents

Runs that last for days. They resume where they stopped.

Each task checkpoints its state, so a failure resumes from the last checkpoint instead of starting over.

Tasks on the critical path, with one failed task resumed from its checkpoint.Illustrative data
Runs across time

Runs across time, as a graph. Some paths only exist across time.

Stacked time slices of one run. Hover a node to follow it.Illustrative

05 · Self-improvement

Agents improve by editing their configuration. Weights stay unchanged.

Failed runs become proposed edits to an agent’s instructions, scored on held-out evals and approved by a person.

A proposed edit, its evals before and after, and the approval.Illustrative data
06 · Model training

Small models, trained for your work. They run on your own hardware.

Narrow steps go to small fine-tuned models that are cheaper and can run on your own hardware.

A small model fine-tuned on task data, with the checkpoint kept before it overfits.Illustrative data
07 · Verification

Checked before it runs. A proof, or the shortest counterexample.

A model checker explores every reachable state and proves the rule holds or returns the shortest sequence that breaks it.

“An entry is posted only with a current approval.” As written, the checker finds submit, approve, edit, post.Illustrative data

Bring us one hard problem. We will show you which of these it needs.

agenthive@qpiai.tech