Your agents run on frontier APIs. They should run on your own models.

We run the post-training loop for your AI agent workloads on infrastructure you control. Distillation cuts inference cost 10 to 20x. Reinforcement learning optimizes the small model on your task. Your data stays in your perimeter.

10 to 20x
inference cost reduction
100%
data stays in your perimeter
4x
faster token generation

Frontier APIs are not a product strategy

Every call to a frontier model sends your data to a third party, locks you into their pricing, and caps your margin. The model gets smarter every month but so does the bill.

💰

Token bills scale with usage

More agents, more calls, more cost. There is no volume discount that beats running a 4B model on your own GPUs.

🔒

Your data leaves your perimeter

Every request, every response, every prompt flows through someone else's infrastructure. For regulated industries, that is a dealbreaker.

🚨

You are locked in

Switching providers means re-engineering prompts and pipelines. The model you depend on can change behavior overnight with no notice.

Four pieces of software. One button that matters.

A drop-in replacement for your frontier API that gets cheaper and smarter every month, plus the machinery that makes that true.

01 / Tracer

Capture production traces

A small library or proxy next to your existing agent. Logs requests and responses into a store you own. Open source. This is the open trace layer that feeds the commercial loop.

02 / Trainer

Turn traces into a small model

Pipeline code that distills teacher traces into a student model, then runs RL to optimize it on the task. Runs as jobs on your GPUs or sovereign cloud. Distillation first, RL second.

03 / Judge

Prove quality with evals

The eval harness scores the small model against the teacher on frozen test sets. Task metrics plus an LLM judge on a 0 to 5 scale. Produces a proof report: quality, cost per task, latency.

04 / Router

Serve with one URL swap

A serving container with an OpenAI-compatible API. Swap one URL. 80 to 90 percent of traffic goes to the small model. The hard 10 to 20 percent goes to a frontier fallback.

On top sits a simple dashboard.

Money saved this month. Quality score. Traffic split. One button: retrain with last month's traces.

That button is the subscription.

The Small Model Distillation Series

Four published blog posts tracking a 0.8B model through progressive distillation stages on a hard agentic benchmark. This is an open learning series, not a product benchmark. Each post breaks down the method, the data, and the honest results.

Blog Post Method Score
Part 1: Offline Teacher-Trace SFT Hard-token SFT on teacher traces 44 / 220 Read →
Part 2: Offline Soft-Label KD Top-k soft-label knowledge distillation 55 / 220 Read →
Part 3: Student-State Correction Teacher correction SFT on student states 67 / 220 Read →
Part 4: On-Policy Probability Distillation OPD with 35B teacher scoring student samples 69 / 220 Read →
Teacher baseline (GPT 5.5 medium) Frontier model, no training 115 / 220

A 0.8B student is 44x smaller than the frontier teacher. The scores show progressive improvement across training stages on a multi-turn SQL tool-use benchmark. Production deployments on narrower, well-defined tasks target 90 to 100 percent of teacher quality at 10 to 20x lower cost.

Send us your traces. Get a cost-cut report in one week.

Free for the first 5 companies. Send 100 to 500 anonymized traces plus your monthly model bill. One week later, you get which steps a small model can take over, the projected cost cut, and the eval plan to prove it.

1. Send traces 100 to 500 anonymized JSONL traces plus your monthly bill
2. We analyze Which steps a small model can take over, projected cost cut
3. You get the report Cost projection, quality plan, eval roadmap. One week.
Request your free audit

No commitment. No sales call. Just the report.

Built by people who ran the loop in public

Not a pitch deck company. The full post-training loop has been published, step by step, with results. The open source trace layer is live and growing.

N

Founder and principal engineer

PhD in multi-agent reinforcement learning. Former CAIO at an acquired AI startup. Former Principal AI Scientist in agentic systems. 10+ years of enterprise ML delivery on AWS and Azure.

The Small Model Distillation blog series (Parts 1 to 4) is the startup's origin story: a 0.8B SQL agent trained through four distillation methods, published openly with benchmark results.