Active Research Infrastructure

Heterogeneous AI
Orchestration for
Production Workflows

A runtime that composes task execution graphs across heterogeneous AI agents. It routes each subtask to whatever agent is currently best at it. Sessions never share context. An output ships only after other agents fail to break it.

Private Research Runtime
Architecture published. Implementation and benchmarks stay private.
Studio Infrastructure
The internal coordination layer of STUDIO HAZE
Real Workloads
Sized against real engineering load at the studio
Architecture-First
Typed DAGs, capability routing, context isolation, validator quorum
System Architecture

Built on first principles

Every layer answers one constraint. Get the most out of heterogeneous inference capacity without letting quality slip.

01 · Agent Registry

Online Capability State

Per-agent n-dimensional capability state is maintained and updated online using task outcomes, validator verdicts, and runtime telemetry. The router selects over this online capability model with no static tags and no manual configuration.

LISPEX
(define (update-state c-a alpha task)
  (let ((r (task-outcome task))
        (v (validator-verdict task))
        (t (runtime-telemetry task)))
    (v+ (vscale (- 1 alpha) c-a)
        (vscale alpha (psi r v t)))))
02 · Task Planner

Typed DAG Decomposition

Requests are decomposed into typed execution DAGs where each node carries I/O schema, QoS constraints, and privacy labels. The planner encodes both data dependencies and parallelizable branches, emitting a fully-typed graph consumed by the scheduler.

LISPEX
(define (make-dag vertices edges types qos)
  (list vertices edges types qos))
03 · Scheduler

Latency-Aware Constrained Dispatch

The scheduler retrieves candidates from an indexed capability store, then applies a constrained assignment over the bounded candidate set, balancing a weighted combination of latency, cost, and risk while satisfying capability thresholds and agent capacity limits.

LISPEX
(define (dispatch tasks agents)
  (letrec ((score
             (lambda (assignment)
               (+ (sum-lat  assignment)
                  (* lambda-c (sum-cost assignment))
                  (* lambda-r (sum-risk assignment)))))
           (feasible?
             (lambda (assignment)
               (and (covers-all?          assignment tasks)
                    (respects-capability? assignment)
                    (respects-capacity?   assignment)))))
    (argmin score
            (filter feasible?
                    (candidate-assignments tasks agents)))))
04 · Context Manager

Minimal-Sufficient Context Projection

Each agent receives the least-privilege context view its subtask requires. Windows are materialized by computing the dependency closure over the session graph, intersecting with per-agent policy filters and subtask schema. Cross-contamination is reduced while preserving the dependency context needed for coherent execution.

LISPEX
(define (context agent subtask)
  (redact
    (intersect
      (closure session-graph
               (ancestors subtask))
      (policy  agent)
      (schema  subtask))))
05 · Validation Layer

Validator Quorum & Adjudication

Critical outputs are routed to an independent validator set. Semantic similarity detects divergence. A composite score over reference grounding, hard constraint checks, and executable validators decides accept or reject. Resolution uses validator quorum rules rather than a simple majority vote.

LISPEX
(define (score output)
  (+ (* w-g (ground   output R))
     (* w-h (hardcheck output C))
     (* w-e (exec     output T))))

(define (accept? outputs)
  (>= (length
        (filter (lambda (o)
                  (>= (score o) tau))
                outputs))
      quorum))
06 · Synthesis Engine

Typed IR Aggregation

Validated outputs are normalized into a canonical typed intermediate representation, then folded through a deterministic synthesis path in topological order along the original DAG. A final schema validator gates the output before delivery.

LISPEX
(define output
  (validate-schema
    (fold-topological
      (map normalize outputs)
      dag)))

Formula snippets are written in Lispex, the deterministic Lisp from the same studio. It runs at lispex.com.

01
Task Input
Structured request + constraints
02
DAG Construction
Dependency analysis + parallelism
03
Agent Dispatch
Capability-matched assignment
04
Parallel Execution
Context-isolated inference
05
Validation
Cross-agent consensus check
06
Synthesis
Schema-conformant output
Capabilities

What this enables

Concrete outcomes made possible by coordinating heterogeneous inference at the infrastructure layer.

Compute Arbitrage

Workloads are routed by a cost-aware policy under SLO constraints. When frontier model pricing shifts, the scheduler follows without manual retuning.

Capability Composition

Specialized agents for code, formal reasoning, and multimodal work compose into a single task workflow. No agent has to be a generalist. Each contributes the thing it is best at.

Fault-Tolerant Execution

Agent failures trigger reassignment through ranked fallbacks, checkpointed state, and idempotent replay. SLA-bound tasks keep moving without duplicated side effects.

Context Efficiency

Each agent gets only what its subtask needs. That keeps per-request cost down and lets context-independent branches run wide.

Auditable Provenance

Every output carries an immutable lineage record, from agent identity and context slice to validator verdict and synthesis path. Debugging and review start from the record, not from memory.

Model-Agnostic Integration

A canonical invocation envelope lets frontier APIs, local runtimes, fine-tuned models, and specialist executors register as first-class agent nodes. No platform changes required.

From the same studio

The studio behind this runs real products

Project Vercel is research infrastructure from STUDIO HAZE, the studio behind Lena Code. The runtime is shaped by what production actually demands. Isolation that holds under deadline. Validation that does not slow the pipeline. Records that survive an audit. Lena Code itself is a separate commercial product and runs on its own proprietary engine.

Lena Codelenacode.com ↗

Lena Code converts source code across languages while keeping its meaning fixed. The proprietary Lena Engine does the deterministic work, rule first. Lena AI steps in where paradigm gaps exceed what any rule set can express.

The Lena Engine holds the deterministic core. Lena AI covers what rules cannot express. The pipeline stays deterministic first, with model help only where it earns its place.

TypeCommercial conversion platform
CoverageMulti-language, M:N
EngineLena Engine · Lena AI
StatusActive deployment
From the same studio

A compiler you can download today

Topaz is a small, closed language for application intent. It reads like Python or TypeScript and compiles to a self-contained native binary with zero crates.io dependencies. STUDIO HAZE ships it at topaz.ooo, versioned and release-signed.

Unicode-first identifiers stay in your language
TOPAZ
function greet(name: string, language: string) -> string {
    return match language {
        case "한국어"  => "안녕하세요, {name}님!"
        case "Русский" => "Привет, {name}!"
        case _         => "Hello, {name}!"
    }
}

let 사용자 = "김토파즈"
print(greet(사용자, "한국어"))
You write the intent. The compiler writes the Rust.
Releasev5.5.0 · download at topaz.ooo
Testscorpus + differential suites
Dependencieszero crates.io
Safetyforbid(unsafe_code)
Outputself-contained native binary
Toolchaintopaz run / build
LicenseApache-2.0
Run traces

Selected traces from internal runs. Workload names and identifiers are removed. What remains is the shape of the coordination. Review rounds, passing checks, and the differential fixtures behind each merged change.

trace AMERGED

A standard-library bridge landed with lazy evaluation and argument order matched to the reference interpreter, case by case.

8 review roundsCI checks passeddifferential fixtures 568 → 586
trace BMERGED

Suggestion diagnostics for unresolved names. Review caught five position-correctness gaps before merge.

5 review roundsCI checks passed
trace CMERGED

An over-reaching static analysis was rejected in review and reshaped into a sound direct-only rule.

CI checks passed
trace DMERGED

Type-alias resolution in typed guards. A suspected soundness hole was investigated and closed with a paired differential fixture.

1 review roundsCI checks passeddifferential fixtures 606 → 615
trace EMERGED

Unknown members on closed union arms are now rejected through one shared predicate.

CI checks passed
trace FMERGED

Multi-module native output pinned against the reference interpreter with a dedicated differential harness.

CI checks passed
Research Direction

At production latency, the bottleneck is coordination, not capability

Raw model capability has scaled dramatically. The dominant unsolved problem is orchestration under constraints. An open-ended task has to break into subtasks an agent can actually be matched to. The pieces run in parallel and come back together at production latency, and no one can afford a manual review on every piece.

Project Vercel is our long-term research infrastructure for this problem. We study online capability calibration, constrained placement under SLO objectives, minimal-sufficient context materialization, and reference-anchored validator quorum. The long-term target is a general-purpose coordination runtime for heterogeneous AI systems.

Capability Measurement
Benchmark-free online calibration of per-agent capability states. Vectors are updated continuously from task outcomes, validator verdicts, and telemetry. No held-out test sets required.
Graph Optimization
Constrained approximation for dynamic subtask placement under latency, cost, and SLO objectives. The current focus is bounded-ratio greedy placement with incremental rebalancing.
Context Compression
Learned minimal-sufficient context views under dependency closure. The goal is the smallest projection that still preserves cross-subtask coherence.
Consensus Mechanisms
Reference-anchored validator quorum and adjudication for distributed inference outputs. Combines semantic divergence detection with hard constraint checks and executable validators.
Deployment Context

Built independently, deployed proudly on Vercel

Project Vercel is an independent research initiative. It is not affiliated with or endorsed by Vercel, Inc. The name is an homage to a way of shipping software we admire. The system itself runs on Vercel-powered deployment primitives.

Independent Project

Project Vercel is an independent initiative by STUDIO HAZE. It is not an official Vercel product, subsidiary, or regional branch.

Built on the Vercel Stack

The system is built around the Vercel deployment platform and benefits from its deployment model, delivery ergonomics, and operational simplicity.

Name with Intent

The name reflects an affinity for the culture of fast iteration, clean deployment surfaces, and composable production systems. It signals admiration for that ecosystem, not formal affiliation.

Vercel is a trademark of Vercel, Inc. Project Vercel is an independent project and is not affiliated with, endorsed by, or sponsored by Vercel, Inc.

Contact

Work with us

We talk with research groups and infrastructure teams that run serious multi-agent workloads. If this work overlaps with yours, write to us.

contact@vercel.co.kr