Respan's banner

Respan

@RespanAI • 1,934 subscribers

Self-driving observability and evals for agents. Formerly Keywords AI (YC W24). Community: https://t.co/JAZ9auu0hr

Shorts

Introducing Span-02: anytime reasoning for decisions. Reasoning models make you pick an effort level and wait. Span-02 takes a deadline and always returns its best decision so far. It gets better the longer it thinks. With reasoning off, it beats Jev at 1/4 of the price, and runs at 1/3 the latency of decision models at its level. It can: - Answer with numbers and scores, not just probabilities and choices - See images and screens (multimodal) - Take a configurable reasoning time And it’s fully open weight!

Introducing Span-02: anytime reasoning for decisions. Reasoning models make you pick an effort level and wait. Span-02 takes a deadline and always returns its best decision so far. It gets better the longer it thinks. With reasoning off, it beats Jev at 1/4 of the price, and runs at 1/3 the latency of decision models at its level. It can: - Answer with numbers and scores, not just probabilities and choices - See images and screens (multimodal) - Take a configurable reasoning time And it’s fully open weight!

221,064 görüntüleme

Introducing Respan Router, powered by Span-01. Automatically routes every request using task context, model capabilities, and cache-aware cost estimates behind 1 model id: 𝐬𝐩𝐚𝐧-𝐫𝐨𝐮𝐭𝐞𝐫 - 13% lower latency than jev-router on τ-bench airline - 37% cheaper than the most accurate single model, at matching accuracy - 0% processing fee on Respan Gateway

Introducing Respan Router, powered by Span-01. Automatically routes every request using task context, model capabilities, and cache-aware cost estimates behind 1 model id: 𝐬𝐩𝐚𝐧-𝐫𝐨𝐮𝐭𝐞𝐫 - 13% lower latency than jev-router on τ-bench airline - 37% cheaper than the most accurate single model, at matching accuracy - 0% processing fee on Respan Gateway

18,112 görüntüleme

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

AI observability platforms raised $1B+ to reinvent print debugging for the agent era. Reading traces manually is not a scalable production workflow. We think the stack should catch issues itself. Meet Respan.

AI observability platforms raised $1B+ to reinvent print debugging for the agent era. Reading traces manually is not a scalable production workflow. We think the stack should catch issues itself. Meet Respan.

1,351,710 görüntüleme • 5 ay önce

Introducing Respan AI Gateway. The world’s first AI gateway with built-in observability, evals, and prompt optimization. Access 1,000+ models through one endpoint, with the production layer teams need beyond routing.

Introducing Respan AI Gateway. The world’s first AI gateway with built-in observability, evals, and prompt optimization. Access 1,000+ models through one endpoint, with the production layer teams need beyond routing.

850,518 görüntüleme • 5 ay önce

Your users never use your product the way you designed it. That’s why we’re launching Prompt Simulations today. Pick a prompt → generate realistic users and scenarios → run full multi-turn conversations → see exactly where your prompt breaks. Test against real behavior before real users find the edge cases.

Your users never use your product the way you designed it. That’s why we’re launching Prompt Simulations today. Pick a prompt → generate realistic users and scenarios → run full multi-turn conversations → see exactly where your prompt breaks. Test against real behavior before real users find the edge cases.

210,543 görüntüleme • 1 ay önce

We told our agent to hack another agent. It found a way in. Then another. Then another. And we can probably hack your agent too. Today, we’re launching Respan Security.

We told our agent to hack another agent. It found a way in. Then another. Then another. And we can probably hack your agent too. Today, we’re launching Respan Security.

202,729 görüntüleme • 3 ay önce

What if you could search your logs by vibe? “The user is pissed” ”The agent is being lazy” “That actually went really well” Introducing Respan Behaviors. 🧵

What if you could search your logs by vibe? “The user is pissed” ”The agent is being lazy” “That actually went really well” Introducing Respan Behaviors. 🧵

140,096 görüntüleme • 3 ay önce

We deserve better error tracking than a giant table. 3 problems shouldn’t be buried in 10,000 error events. Introducing Respan Error Tracking.

We deserve better error tracking than a giant table. 3 problems shouldn’t be buried in 10,000 error events. Introducing Respan Error Tracking.

127,227 görüntüleme • 3 ay önce

Introducing Respan P-1, the world’s best privacy model. Every AI application handles sensitive data. Existing approaches to protecting it force a tradeoff between accuracy, latency, and cost. We built P-1 to eliminate that tradeoff. P-1 sets a new state of the art for detecting and redacting sensitive information - outperforming existing models across our benchmark while being fast enough to run on every request. Starting today, P-1 powers PII redaction across Respan.

Introducing Respan P-1, the world’s best privacy model. Every AI application handles sensitive data. Existing approaches to protecting it force a tradeoff between accuracy, latency, and cost. We built P-1 to eliminate that tradeoff. P-1 sets a new state of the art for detecting and redacting sensitive information - outperforming existing models across our benchmark while being fast enough to run on every request. Starting today, P-1 powers PII redaction across Respan.

73,992 görüntüleme • 2 ay önce

Introducing Respan Evals 2.0: Automation. Most AI production issues don’t show up as errors. They show up as answers that look valid but are wrong, incomplete, unsafe, or off-brand. Respan now runs automated online evals on live production logs to catch silent quality failures in real time.

110,832 görüntüleme • 4 ay önce

Most AI products don’t fail. They just lose a little performance everywhere. We built Respan for that entire loop, and that’s the idea behind our new film.

Most AI products don’t fail. They just lose a little performance everywhere. We built Respan for that entire loop, and that’s the idea behind our new film.

70,988 görüntüleme • 2 ay önce

Finta uses AI across its accounting platform for startups, with one rule: accuracy over automation. Every day, Finta uses Respan to test prompts across models, manage golden datasets, and run evals before shipping. See how Andy and his team build reliable AI accounting ↓

Finta uses AI across its accounting platform for startups, with one rule: accuracy over automation. Every day, Finta uses Respan to test prompts across models, manage golden datasets, and run evals before shipping. See how Andy and his team build reliable AI accounting ↓

54,670 görüntüleme • 1 ay önce

Respan Launch Week, Day 2: Evals Evals are one of the hardest parts of building AI applications. It is not because teams cannot run them. It is because they are hard to structure, hard to compare, and even harder to improve over time. So teams end up guessing. Did this prompt actually get better? Is this model really an improvement? We built Evals in Respan to make this systematic. You define: - what you want to test, such as prompts, models, or configs - the dataset, from production logs or test cases - the evaluators, whether LLM, code, human, or a mix Then you run experiments and compare results side by side. Same inputs. Same evaluation. Clear answers. No more guessing. Start running your first eval.

Respan Launch Week, Day 2: Evals Evals are one of the hardest parts of building AI applications. It is not because teams cannot run them. It is because they are hard to structure, hard to compare, and even harder to improve over time. So teams end up guessing. Did this prompt actually get better? Is this model really an improvement? We built Evals in Respan to make this systematic. You define: - what you want to test, such as prompts, models, or configs - the dataset, from production logs or test cases - the evaluators, whether LLM, code, human, or a mix Then you run experiments and compare results side by side. Same inputs. Same evaluation. Clear answers. No more guessing. Start running your first eval.

127,680 görüntüleme • 6 ay önce

We worked with OpenAI to build a native integration for the OpenAI Agents SDK. Today, we are very excited to launch as a tracing processor. With just a few lines of code, you can trace all your agent workflows and debug them much faster. Check out this quick demo to get started:

We worked with OpenAI to build a native integration for the OpenAI Agents SDK. Today, we are very excited to launch as a tracing processor. With just a few lines of code, you can trace all your agent workflows and debug them much faster. Check out this quick demo to get started:

58,186 görüntüleme • 1 yıl önce

Daha fazla içerik yok.