Aman Pandey

251 posts

Aman Pandey banner
Aman Pandey

Aman Pandey

@ixchio

17 PRs merged into Gemini CLI, Astro, mem0, OpenHands. Built agent-vcr: time-travel debugger for AI agents. Founding Member, OpenHands Champions (10 globally)

India Katılım Ağustos 2022
97 Takip Edilen35 Takipçiler
Aman Pandey
Aman Pandey@ixchio·
@yabsssai AI changes its answer the moment you push back.. built mindfuck so it can't.
English
0
0
0
4
YabsssAI
YabsssAI@yabsssai·
@ixchio grateful for that openness and willingness to share your product on product hunt, it really takes a lot of courage to put something like blurt out there for feedback - what motivated you to build blurt and what problem were you trying to solve?
English
1
0
0
12
Aman Pandey
Aman Pandey@ixchio·
shipped something today every AI tool agrees with you. tells you your idea is great. maybe needs some refinement. very cool very exciting i got tired of it so i built mindfuck — it tells you where your idea breaks, what problem you're actually solving, and what to do about it. monday morning. concrete. no vibes pip install mindfuck (yes that's the real name yes it's on pypi...yes i checked)
Aman Pandey tweet media
English
1
0
0
48
Aman Pandey
Aman Pandey@ixchio·
@tsenart built agentic infra handling 1200+ daily executions found two CVEs...shipped across 65k-star OSS maybe I'm your unicorn
English
1
0
1
323
Tomás Senart
Tomás Senart@tsenart·
Starting to map the graph for a founding engineer at Perfloop. Looking for real systems depth - perf, infra, observability, AI infra, devtools, databases, runtimes - plus extreme agency, product/architecture judgment, and native agentic building. Europe / near-Europe overlap preferred.
English
7
6
34
6K
Theofanis Karaletsos
Theofanis Karaletsos@Tkaraletsos·
Hiring AI/ML talent at @achira_ai to build world models for molecules We're looking for scientists and research engineers who will join Achira's technical staff to join its mission to build microscopic world models. As large scale AI models have matured and revolutionized our ability to intervene on the digital world, we are increasingly racing towards physical AI as the next frontier. Achira builds foundation simulation models that will explictly model the microcosm on an atomistic level with a goal to deeply impact virtual chemistry in service of drug discovery and is marrying the best of deep learning/generative AI with physics, statistical mechanics, and quantum chemistry. If you're up for the challenge to build at the frontier, come join us. achira.ai/careers/
English
11
35
388
153.5K
Aman Pandey
Aman Pandey@ixchio·
@jacobeffron @swyx this. i run 1,200+ agent executions/day and "stability" just means you've seen every failure mode enough times to have a handler for it
English
0
0
0
25
Jacob Effron
Jacob Effron@jacobeffron·
.@swyx on whether AI infrastructure has finally stabilized:
Jacob Effron@jacobeffron

Always enjoy getting to chat with @swyx on our annual cross-episode with @latentspacepod on the state of AI. We hit on what’s shifted, what surprised us and what’s next. We covered: ▪️ Whether AI infrastructure has finally stabilized ▪️ Implications of agents buying developer tools ▪️ The AI coding wars ▪️ The foundation model vibe shift ▪️ Why Swyx reversed his view on open models ▪️ When to train your own model ▪️ What's top of mind for the best AI engineers YouTube: youtu.be/A_7WafI9bhE Spotify: bit.ly/3QHcCix Apple: bit.ly/4eJERaa 0:00 Intro 1:17 What the Top AI Engineers Are Thinking About 2:13 Has AI Infra Finally Stabilized? 6:39 When Does Doing RL In-House Make Sense? 11:26 Why Selling Dev Tools to Agents is Different 17:18 AI Coding Wars 29:04 Consumer AI Plateau 30:22 Codex vs Claude Code 44:52 Future of Open Models

English
4
6
28
12.4K
Aman Pandey
Aman Pandey@ixchio·
@MelkeyDev agent infra meetup and i'm sitting in India with 3 PRs open in agent tooling watching this from my feed at 1am 😭 someone remote me in
English
0
0
1
51
Aman Pandey
Aman Pandey@ixchio·
@nizzyabi ironic timing .... i have 3 PRs open in Gemini CLI just sitting there waiting for review lol
English
0
0
0
89
Aman Pandey
Aman Pandey@ixchio·
@katedeyneka @ycombinator Reelful's camera-roll-to-reel flow is exactly what agentic pipelines should feel like. 👏 congrats on YC app .... rooting for the W
English
0
0
0
12
Aman Pandey
Aman Pandey@ixchio·
Name one thing that sounds fake but is actually true about building AI agents in production. I'll go first: the hardest part isn't the AI, it's the retry logic.
English
0
0
0
74
Aman Pandey
Aman Pandey@ixchio·
Just dug into @EntelligenceAI adversarial verification loop...most AI code reviewers: flag everything suspicious → noise → engineers stop readingEllie: generate finding → actively try to disprove it → only surfaces if it survives That's not a feature. That's a product philosophy. Rare.
English
0
0
1
66
Aman Pandey
Aman Pandey@ixchio·
@sorenrood @runcomputing the hardest part of this role isn't the agent architecture it's that ops workflows were never designed to be observed. no logs, no states, no handoffs. you're basically building the observability layer before you can build the automation.
English
0
0
0
296
søren
søren@sorenrood·
we are hiring a couple FDEs at @runcomputing. you’ll travel around the country, sit down with people that run operationally heavy businesses, and help them deploy background agents that own end-to-end operational work. ideal background: you’ve worked close to real operations: manufacturing, distribution, supply chain, consulting, or hands-on implementation. you can move between the shop floor, exec room, and eng team without losing context. comfortable with ambiguity. able to turn it into something that actually works for customers. email / dm me.
English
13
4
117
10.8K
Aman Pandey
Aman Pandey@ixchio·
@bnafOg @OpenHandsDev @mem0ai stateless by design each run is fully isolated zero persistence. IndexedDB is the natural next layer but I wanted the no backend proof solid before adding state complexity.
English
0
0
0
5
Bnaf.OG | 🟧
Bnaf.OG | 🟧@bnafOg·
@byteakp @OpenHandsDev @mem0ai Zero-backend browser-native is a real architectural bet — great for privacy and cold-start latency. Curious how you handle agent state persistence across sessions. IndexedDB? Or does the workstation treat each run as stateless by design?
English
1
0
1
26
Aman Pandey
Aman Pandey@ixchio·
@nicoloboschi @mem0ai interesting the bug I found was specifically in the BM25 lemmatization path on the async side, which made scores always return zero curious how Hindsight handles that in the recall pipeline. gonna dig into the repo...
English
0
0
0
10
Aman Pandey
Aman Pandey@ixchio·
found a silent bug in @mem0ai that zeroed out BM25 keyword search for all async users the kind that raises no errors and breaks everything PR merged ✓
Aman Pandey tweet media
English
1
0
5
104
Aman Pandey
Aman Pandey@ixchio·
@gneubig the clarification-performance tradeoff hits different when you've watched agents spiral on underspecified tasks in the wild fewer, smarter questions > more dumb ones every time makes me think about how OpenHands handles ambiguous instructions at the EventStream level.
English
0
0
0
72
Graham Neubig
Graham Neubig@gneubig·
In this work, we examined the question: "what makes a good clarifying question?" in the context of software engineering. We trained a model that was specifically tasked with asking clarifying questions, allowing for improved results with fewer questions.
Sanidhya Vijayvargiya@sanidhya903

1/ Humans often can’t state exactly what they want, making things hard for AI agents. Obvious fix: ask clarifying questions. But which ones?  We studied this empirically with coding agents. Effective clarification comes down to two properties: answerability and task relevance.

English
1
6
41
7.3K

Keşfet

@yabsssai @tsenart @Tkaraletsos @achira_ai @jacobeffron @swyx @MelkeyDev @nizzyabi