DEV Community: Kumar Abhishek

Jev — AI That Decides , Not Talks

Kumar Abhishek — Sat, 26 Sep 2026 12:53:56 +0000

What is Jev?? 🤔

Is it an LLM?

Is it another Frontier Reasoning Model?

No.

Then… what exactly is it?

This week, one of the more interesting AI launches caught my attention:

Jev — TypeSafe AI’s first “System One” model.

And I think the interesting part is not that it is “another AI model”.

The interesting part is what it is designed NOT to do.

What is Jev?

Jev is a decision model, designed to make fast, structured decisions that software can consume directly.

Traditional LLMs are generally autoregressive:

Prompt → Token → Token → Token → … → Response

They generate a sequence of text, one token at a time.

Jev takes a very different approach:

State + Questions → Structured decisions

Instead of generating an explanation, paragraph or code, Jev evaluates predefined questions and returns typed answers with probabilities and confidence.

Think:

LLMs generate words for humans.
Jev generates decisions for software.

TypeSafe calls this a System One model — a model optimized for fast, machine-consumable decisions rather than open-ended text generation.

How does Jev work?

Imagine I give the model an unstructured piece of business state:

📩 A customer support ticket
📊 A transaction record
🖥️ A system log
🤖 An agent trace

…and I define several questions:

1. Choice
“What is the category?”

→ Billing: 0.10
→ Technical: 0.82
→ Sales: 0.08

2. Score
“How urgent is this?”

→ Low / Medium / High + score/distribution

3. Noul
“Is this ticket urgent?”

→ Probability = 0.97

And here is the interesting part:

You can ask multiple questions about the same state in one request.

Jev evaluates those decisions in parallel rather than generating a long sequence of tokens for each answer.

So instead of:

“First let me explain what I think…”

You get:

Here are the decisions your software needs.

And Jev does NOT generate text.

This is probably the biggest conceptual shift.

Jev is not designed to write:

❌ emails
❌ essays
❌ code
❌ explanations
❌ chat responses

Its output space is predefined and typed.

That means software does not need to parse a paragraph and hope that the model followed the requested JSON format.

The interface itself is structured.

But one important nuance:

Type-safe output does NOT mean the model can never be wrong.

Jev can still make an incorrect judgment.

The advantage is that the system gets a structured decision plus uncertainty/probability, rather than an unconstrained block of generated text.

The numbers are also interesting 👀

According to TypeSafe AI, Jev is targeting:

⚡ 70–500 ms end-to-end latency

💰 $0.042 / 1M input tokens

🪶 No metered output-token cost

And TypeSafe attributes this to its different architecture, parallel sampling and its training approach called:

RLCD — Reinforcement Learning for Calibrated Decisions.

These are vendor-reported figures, so I’d treat them as launch claims rather than universal benchmarks.

So… where would I actually use Jev?

This is where it gets interesting.

I would not think:

“Jev will replace my LLM.”

I would think:

“Where in my software am I repeatedly asking AI to make a small bounded decision?”

For example:

1️⃣ Model Routing

Not every request needs your most expensive reasoning model.

You could do:

User Request → Jev → classify complexity → choose model

Simple lookup
→ Fast model

Complex debugging
→ Frontier reasoning model

This lets AI decide which AI should do the work. LangChain has already demonstrated Jev-based model-routing middleware.

2️⃣ Agent Safety / Tool Gating

Imagine an agent wants to execute:

delete_file()

send_email()

run_bash()

deploy_production()

Before the tool executes:

Tool Request → Jev → Risk Decision → Policy → Allow / Block / Human Review

This is particularly interesting for agentic systems because the decision layer can sit directly inside the agent loop.

LangChain has demonstrated this pattern with Jev-based Auto Mode middleware for checking risky tool calls before execution.

3️⃣ Real-time Classification & Routing

Think:

Customer support
Fraud triage
Incident management
Priority detection
Content moderation
Lead routing
Transaction checks
Operational alerts

These systems often don't need an essay.

They need:

“Which bucket?”

“How urgent?”

“Should I escalate?”

“Is this safe?”

“Which workflow should run next?”

That's exactly where a decision-oriented model becomes interesting.

When should you use Jev?

I would look for four characteristics:

The decision is bounded.
You already know the possible outcomes.

The decision happens frequently.
Thousands or millions of decisions make latency and cost meaningful.

The application needs the decision, not the explanation.

The output can directly drive software logic.

For example:

if urgent_probability > threshold → escalate

if category == "technical" → route_to_engineering

if risk_probability > threshold → block_tool

That is a very different role from:

“Explain to me why this happened.”

A simple business architecture 👇

Customer / System Event
↓
Business State + Context
↓

⚡ Jev

Urgent?
Category?
Risk?
Priority?
Which model?
↓

Policy / Business Rules

Thresholds + guardrails + permissions
↓
Fast Path → Automated Action

Complex Case → Frontier LLM

Uncertain Case → Human Review

This, to me, is where the idea becomes much more powerful.

Jev doesn't necessarily replace the reasoning model.

It can sit before, between, and after reasoning models.

And this is where my biggest takeaway starts:

The probabilities need policy around them.

A probability by itself isn't an action.

0.87 urgent

doesn't tell the system what to do.

Your policy does.

0.87 → escalate

0.62 → ask another question

0.31 → continue normally

And that leads to another important engineering question:

Observability cannot stop at “what did the model predict?”

We need to observe:

Decision → Probability → Policy → Action → Outcome

Because a highly confident wrong decision can be much more dangerous than an uncertain one.

And that's perhaps the most interesting part of the Jev conversation for me.

Jev does not belong everywhere. 😛

It belongs where AI needs to make many small, fast, structured decisions inside software.

So rather than asking:

“Can Jev replace my LLM?”

I think the better question is:

“Which decisions in my application are currently being made by an expensive text-generating model… when all I really needed was a fast decision?”

Next post → How to actually use Jev: API, typed questions, architecture and a real workflow example. 🚀

AI #GenerativeAI #AgenticAI #AIEngineering #LLM #SystemOne #Jev #TypeSafeAI @typesafe https://typesafe.ai/

Navigating the NVIDIA Tech Ecosystem

Kumar Abhishek — Thu, 01 Jan 2026 11:40:24 +0000

The Complete Roadmap to Immersing Yourself in the NVIDIA Tech Ecosystem: From Curiosity to Mastery

The tech world is currently undergoing a seismic shift, and at the epicenter of this revolution is one name: NVIDIA. While many know them for gaming GPUs, the reality is that NVIDIA has built a massive, multi-layered ecosystem that powers everything from Generative AI and digital twins to autonomous vehicles and quantum computing.

If you are looking to start something productive in tech, immersing yourself in the NVIDIA ecosystem isn't just a hobby—it’s a career-defining move.

Here is your step-by-step guide to navigating and mastering the NVIDIA landscape.

Phase 1: The Discovery Portals (Mapping the Terrain)
Before you write a single line of code, you need to understand the "What" and the "Why." NVIDIA’s technology is vast, covering hardware architecture, software libraries, and industry-specific frameworks.

The Big Picture: Start at the Official Website to understand the company’s current focus.
The Cutting Edge: Visit the NVIDIA Research Portal. This is where you see the future of AI, computer vision, and graphics before it hits the mainstream.
The Tech Stack: Explore NVIDIA Technologies to learn about the architectures (like Blackwell or Hopper) and the industry-specific breakthroughs they are driving.

Phase 2: Learning and Certification (Building the Foundation)
Once you’ve mapped the terrain, it’s time to gain the skills. NVIDIA offers some of the most rigorous and respected training in the industry.

NVIDIA Learn: Your starting point for educational resources is the Learn Page.

The DLI (Deep Learning Institute): This is the crown jewel of NVIDIA’s training. The DLI offers self-paced and instructor-led courses. Many are free and provide hands-on access to fully configured GPU servers in the cloud.

Get Certified: To prove your expertise, look into NVIDIA Certifications. They are categorized into two logical paths:

Infrastructure Professional: Focuses on hardware, networking (InfiniBand), and data center setup.
Developer: Focuses on software, AI models, and programming (CUDA, Omniverse).

Phase 3: The Professional & Corporate Edge
If you are part of a startup or a large enterprise, NVIDIA provides specific entry points to help you scale your products.

NVIDIA Connect: If you are a software developer or service provider, join the NVIDIA Connect Program. It’s a free program designed to help you shorten your time-to-market with technical guidance and preferred pricing.

Corporate Partner Portal: Check if your company is already an NVIDIA partner at the Partner Portal. Registering at nvid.nvidia.com can grant you additional access to exclusive courses and discounts on certifications.

NVIDIA Academy: For those focused on the "how-to" of enterprise systems, the NVIDIA Academy is a goldmine for learning about DGX systems, networking (Cumulus Linux, RDMA), and data science tools.

Phase 4: The "Doer’s" Paradise (Hands-On Development)
This is where the rubber meets the road. If you want to get your hands dirty, these are the platforms you need to bookmark.

NVIDIA Documentation Hub: The Docs Hub is the "Bible" for any NVIDIA developer. It contains the technical manuals for every tool and SDK NVIDIA offers.

The Developer Portal: Your home base. Register for the NVIDIA Developer Program. Once registered, you can:

Connect your GitLab account for seamless version control.

Join the Google Cloud & NVIDIA Community to leverage cloud-based GPU power.

Phase 5: Build and Deploy (The AI Factory)
The most exciting part of the ecosystem right now is build.nvidia.com. This is NVIDIA’s portal for NIMs (NVIDIA Inference Microservices) and hosted APIs.

Why use it?

Free Serverless APIs: You can test state-of-the-art models (like Llama 3, Mistral, or NVIDIA’s own Nemotron) directly through your browser.

Validated Blueprints: Don't start from scratch. Use their blueprints to build "Agentic AI" or RAG (Retrieval-Augmented Generation) applications.

NVIDIA Cloud Services: Once you register at build.nvidia.com, you are automatically allocated an ID on cloudaccounts.nvidia.com, linking you to their vast cloud infrastructure.

Final Thoughts
Immersing yourself in the NVIDIA ecosystem is no longer just about "coding for GPUs." It’s about understanding how to orchestrate the most powerful computing stack ever created. Whether you are an individual developer, a student, or a tech lead, the resources are there—most of them for free.

The best time to start was yesterday. The second best time is now.

Let’s Navigate the NVIDIA Frontier Together

Learning in public is the best way to master a stack as complex as NVIDIA’s. Over the coming weeks, I’ll be tackling a series of DLI courses and sharing “Cheat Sheets” for those looking to follow in my footsteps.

I’m currently at a crossroads between two fascinating paths:

The World of Brains: Generative AI and Agentic AI (Building autonomous, reasoning agents with NIMs)
The World of Bodies: Physical AI and Robotics (Exploring Isaac Sim and the frontier of embodied AI)

If you are a developer, architect, or tech enthusiast, hit that follow button. Let’s exchange insights as we move from “learning” to “building.

Which part of the NVIDIA ecosystem are you most excited to explore? Let me know in the comments!

NVIDIA #AI #DeepLearning #TechCareer #AI_SoftwareDevelopment #GPUComputing #GenerativeAI #Nemotron