Kalilur Rahman (@krahman) header
Kalilur Rahman (@krahman) profile picture

Tech Savvy Professional, Interested in general knowledge sharing and lover of good things. RTs not endorsements. Use automation a lot.

India
Joined May 2007
Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)Photo by Kalilur Rahman (@krahman)
Python Data Visualization Essentials Guide ISBN: 9789391030070 Build your data science skills. Start data visualization Using Python. Right away. Become a good data analyst by creating quality data visualizations using Python.  #python #datavisualization #ArtificialIntelligence
2
1
273
Kalilur Rahman retweeted
andrej karpathy: “the future belongs to people who turn ideas about LLMs, RAG, and AI agents into running systems” this free AI Engineering Hub does exactly that — 93+ production-ready projects on LLMs, RAG, MCP and AI agents this is the clean path i wish i had earlier 01 AI Engineering Roadmap github.com/patchy631/ai-engi… 02 LaTeX OCR with Llama github.com/patchy631/ai-engi… 03 Local ChatGPT with DeepSeek github.com/patchy631/ai-engi… 04 Fastest RAG Stack github.com/patchy631/ai-engi… 05 Agentic RAG with DeepSeek github.com/patchy631/ai-engi… 06 Multi-Agent Deep Researcher github.com/patchy631/ai-engi… 07 DeepSeek Fine-tuning github.com/patchy631/ai-engi… 08 Real-time Voice Bot github.com/patchy631/ai-engi… 09 Deploy Agentic RAG github.com/patchy631/ai-engi… 10 Evaluation and Observability github.com/patchy631/ai-engi… 11 Qwen3 vs DeepSeek-R1 github.com/patchy631/ai-engi… 12 MCP Agentic RAG with FireCrawl github.com/patchy631/ai-engi… 13 NotebookLM Clone github.com/patchy631/ai-engi… 14 Financial Analyst with DeepSeek github.com/patchy631/ai-engi… 15 The full hub. Free. github.com/patchy631/ai-engi… you either open this now or keep collecting random playlists for another year save this ↓
7
55
1
301
12,417
Kalilur Rahman retweeted
This is getting f*cking scary. 15 GitHub repos now wrap REA in a desktop studio, a local-model plugin and a scanner that shows what it could pull out of your own app [01] wrappers and installs ERA: wraps REA with run records, reports and a desktop studio github.com/dav1dngy/era-agen… reverseport: turns findings into rebuild specs and ports between engines and languages github.com/laaielichter/reve… rea-mcp: REA inside LM Studio for local models github.com/Nowng/rea-mcp rea-flake: REA with Ghidra packaged for Nix github.com/Fractal-Tess/rea-… [02] check and defend reaproof: scans your own build for what REA could recover, like source maps and keys github.com/Junhan2/reaproof antirea: stop directives that tell agents not to run REA on your project github.com/MockirDev/antirea REA-demo: REA analyzing its own package, graded against the source github.com/az9713/REA-demo [03] worked examples OpenBSD_netcat_REA: a static analysis of netcat-openbsd, every claim linked to an evidence file github.com/NullLabTests/Open… hexgrab: a color picker written from REA’s study of PowerToys behavior, with no PowerToys code github.com/iammukeshm/hexgra… [04] reading rea-research: a deep write-up of REA’s architecture and trust model, in Chinese github.com/ringozzt/rea-rese… reatool: unofficial install guides and a comparison with GhidraMCP and ida-pro-mcp github.com/ZooHero500/reatoo… [05] maps Awesome-REA: a bilingual catalogue of REA tools, cases and papers github.com/reacher-z/Awesome… awesome-rea: every entry names its own acceptance test github.com/yibie/awesome-rea rea-case-rankings: cases from GitHub, HN, X and Reddit on a 1 to 5 score github.com/miles990/rea-case… rea-gems: an auto-updated gallery of REA cases github.com/sqwu/rea-gems how the loop runs: app -> REA -> findings -> spec -> rebuild -> check -> harden start with an app you own and watch it explain itself.
9
110
4
649
42,379
Kalilur Rahman retweeted
🚨 GitHub تحلّ رسمياً أكبر مشكلة في الـ Vibe Coding! أطلقت GitHub أداة جديدة باسم Spec Kit، وتجاوزت 120 ألف نجمة على GitHub في وقت قياسي جداً! 🔥🧵 بدلاً من كتابة برومبت عشوائي والدعاء ألا يدمر الذكاء الاصطناعي مشروعك.. Spec Kit تجبر الـ AI على كتابة مواصفات وهيكلة منتظمة قبل لمس سطر كود واحد! كيف يعمل النموذج؟ 1️⃣ يفهم ما تريد بناءه بدقة 2️⃣ يطرح عليك الأسئلة والمعطيات الناقصة 3️⃣ يهيكل المشروع بالكامل 4️⃣ ثم يبدأ في كتابة الكود! النتيجة؟ ساعات أقل في إصلاح الأخطاء العشوائية، كود نظيف ومنظم، ونتائج مضمونة عند التعامل مع الوكلاء الذكيين (Agents). 🔄 سلسلة أوامر العمل بسيطة ومباشرة: • constitution/ 👈 القواعد والمعايير الأساسية • specify/ 👈 ما الذي تريد بناءه بالضبط؟ • clarify/ 👈 أسئلة وتوضيحات قبل البدء • plan/ 👈 المعمارية والتقنيات المستخدمة (Stack) • tasks/ 👈 تقسيم المهام بالترتيب • implement/ 👈 مرحلة التنفيذ والتكويد 💡 المفاجأة: الأداة متوافقة ومجربة مع أكثر من 25 وكيل ذكي (مثل Cursor, Claude Code, Copilot, Gemini CLI, Codex وغيرهم). • 🌟 أكثر من 120,000 Star • 🍴 أكثر من 10,000 Fork 🔓 مفتوحة المصدر بالكامل من GitHub 📌 احفظ التغريدة للرجوع إليها وقت الحاجة 🔖 رابط المستودع (Repo) تجده في أول تعليق 👇
36
305
4
2,680
242,412
Kalilur Rahman retweeted
If I had to land a $200K AI engineer job in 90 days, I would not get a degree. I would master these 10 GitHub repos. 1. awesome-llm-apps The production AI playbook. RAG, agents, multimodal apps, all in working code. 106K+ stars. Repo → github.com/Shubhamsaboo/a… 2. LangChain The foundational framework. Used in production by Klarna, Replit, Elastic, and most AI startups in 2026. Repo → github.com/langchain-ai/l… 3. LangGraph The orchestration layer powering production agents. The skill on every senior AI engineer job description. Repo → github.com/langchain-ai/l… 4. CrewAI Multi-agent coordination. The framework most Fortune 500 teams reach for first. Repo → github.com/crewAIInc/crew… 5. Ollama Run any open-source LLM on your own machine. The fastest way to learn how models actually work. Repo → github.com/ollama/ollama 6. awesome-mcp-servers MCP is the standard every major AI lab adopted in 2026. Knowing it puts you ahead of 99% of engineers. Repo → github.com/punkpeye/aweso… 7. Qdrant The vector database used for production RAG at scale. Embeddings and semantic search are non-negotiable for AI roles. Repo → github.com/qdrant/qdrant 8. AI-Agents-for-Beginners Microsoft's free 12-lesson course on building agents. Real code, real exercises, real prep. Repo → github.com/microsoft/ai-a… 9. system-design-primer Production AI is system design. The repo FAANG engineers use to prep for interviews. Repo → github.com/donnemartin/sy… 10. awesome-claude-code The playbook for the tool now used inside FAANG, OpenAI, Anthropic, and most YC startups. Repo → github.com/hesreallyhim/a… Here's the wildest part: A $200K AI engineer in 2026 isn't paid for a degree. They are paid for what these 10 repos teach. The market doesn't care where you learned it. It only cares if you can ship. 90 days. 10 repos. One portfolio that proves you can do the work. That's it. That's the whole game. Save this before you forget. 100% free. 100% open source.
30
153
7
964
47,968
Kalilur Rahman retweeted
holy f*ck, web designers have NO IDEA what’s coming after them in 2027 the ones learning AI tools right now are getting ahead. the rest might be fighting for jobs they used to take for granted. 7 GitHub repos every designer should know before it’s too late: 1. Graphify 137k stars and in Y Combinator. point it at your old projects, brand PDFs and screenshots, and Claude gets a memory of everything you've ever designed. "what did this client hate last time?" answered in seconds github.com/Graphify-Labs/gra… 2. Awesome DESIGN.md 84 design systems pulled from real sites, each one a single markdown file. drop it in your project, say "build me a page that looks like this", done github.com/VoltAgent/awesome… 3. Hallmark a design skill from Together AI that refuses to look AI-generated. 50 themes, checks every page in a real browser, and /hallmark redesign rebuilds any old page github.com/Nutlope/hallmark 4. Huashu Design one sentence in, a launch animation, clickable app prototype, editable PPT or infographic out. it shows you 6 directions first, you pick 3 github.com/alchaincyf/huashu… 5. Designer Skills Pack 273 skills and 76 commands for Claude Code: research, design systems, UI, interaction. a whole design department in one install github.com/Owl-Listener/desi… 6. Motion Design Skill LottieFiles taught AI to think like a motion director: timing, easing, choreography, Disney's 12 principles adapted for UI. this is where the "feel" comes from github.com/LottieFiles/motio… 7. Blender Agent Skills 22 pro Blender workflows for Claude Code: modeling, materials, rigging, animation, rendering. the full 3D pipeline, not just "make a cube" github.com/EL4CTEO/blender-s… the cell you're watching? built solo with Claude Code plus a few of these tools. no agency, no team, no After Effects. want to make something like this? like + save this post, comment "CELL" or DM me and I'll send you the exact prompt.
132
243
18
1,487
118,803
Kalilur Rahman retweeted
天哪,Karpathy 在 GitHub 上免费放了一个完整的 AI 学位课程 7 个仓库,306,741 星标,而且按正确顺序学习,它们会带你从反向传播到能通宵运行研究的智能体 。 每个仓库都是前一个的下一层。仓库地址👇
37
318
12
1,699
100,388
Kalilur Rahman retweeted
if you want to build an AI trading system from scratch, start with this article you will end up with real market data, real backtests and a bot that trades for you 24/7
187
242
86
1,716
2,691,208
Kalilur Rahman retweeted
this Grok Bot terminal claims it turned $14.82 into $7,643.21 overnight wild number. but the six-agent setup is the part worth looking at instead of asking one chatbot what to buy, the author split the workflow: → scanner finds new pairs → volume agent flags unusual activity → safety agent checks for rug risks → sizing agent decides position size → execution agent handles orders → logging agent records decisions and their reasons the terminal puts the whole process on one screen an equity curve next to the activity feed enter / skip / wait / exit decisions risk checks before execution entry and exit prices with slippage win and loss stats updating as trades close the author reports: starting balance: $14.82 ending balance: $7,643.21 win rate: 68.3% logged events: 1,412 those are self-reported figures, not an independently verified trading record and 68.3% wins doesn’t settle the question. small wins can still get wiped out by a few oversized losses what makes the architecture interesting is being able to trace a trade backward what triggered it which checks passed why that size was chosen what price actually filled what caused the exit that’s much more useful than a green balance with no history behind it if you’re building something similar, make the decisions inspectable before making the profit screenshot impressive save this for the agent split, not the overnight return
126
89
5
1,058
288,910
Kalilur Rahman retweeted
THIS IS INSANE Anthropic just dropped a goldmine of 20 Claude Code plugins. $29,077 worth of tools, FREE on GitHub paid marketplaces sell a skill for $19 and an agent for $39. these repos ship 970 skills and 273 agents for $0 1.8M stars combined, picked from 68,900 repos tagged claude-code most people still run Claude Code bare: no memory, no team, no taste, burning tokens on raw logs 20 GitHub projects you should open first: 1. Everything Claude Code → 71 agents, 311 skills and 97 commands in one harness. the closest thing to an agent OS github.com/affaan-m/everythi… 2. agent-skills → 25 production skills by Addy Osmani: spec, plan, build, verify, review, ship github.com/addyosmani/agent-… 3. Ruflo → runs agent swarms across Claude Code and Codex with shared memory between them github.com/ruvnet/ruflo 4. agents → 202 agents, 185 skills and 105 commands packed into 94 installable plugins github.com/wshobson/agents 5. claude-mem → records what your agent did, summarizes it and feeds it back next session github.com/thedotmack/claude… 6. planning-with-files → keeps the plan in markdown on disk, so it survives /clear, compaction and crashes github.com/OthmanAdi/plannin… 7. Graphify → turns code, docs and SQL schemas into a knowledge graph. parsed locally, nothing leaves your machine github.com/Graphify-Labs/gra… 8. Understand-Anything → an interactive map of your codebase with search, guided tours and diff impact analysis github.com/Egonex-AI/Underst… 9. ponytail → makes Claude code like the laziest senior dev. 53% less code and 45% fewer tokens in its own Opus 5.5 benchmark github.com/DietrichGebert/po… 10. rtk → a Rust CLI proxy that cuts 60-90% of the tokens from common dev command output github.com/rtk-ai/rtk 11. headroom → compresses logs, tool outputs and JSON before they reach the model github.com/headroomlabs-ai/h… 12. context-mode → sandboxes tool output. 315 KB in, 5.4 KB to the model github.com/mksglu/context-mo… 13. UI UX Pro Max → 79 styles, 192 palettes and 74 font pairings, so your UI stops looking AI-made github.com/nextlevelbuilder/… 14. taste-skill → gives Claude art direction, real reference photos and a browser check at several screen widths github.com/Leonxlnx/taste-sk… 15. video-shotcraft → turns your real pages into a cinematic product video in Remotion, synced to the beat github.com/Vincentwei1021/vi… 16. Agent-Reach → lets your agent read YouTube, GitHub, Reddit and the web from one CLI, zero API fees github.com/Panniantong/Agent… 17. cc-switch → a desktop app to switch providers in one click and manage MCP, skills and prompts in one place github.com/farion1231/cc-swi… 18. pm-skills → 69 product skills and 42 chained workflows, from discovery to launch github.com/phuryn/pm-skills 19. claude-skills → 380+ skills across engineering, marketing, finance, compliance and C-level advice github.com/alirezarezvani/cl… 20. career-ops → scans job boards, scores every role against your CV and drafts the application. you still press Submit github.com/career-ops-hq/car… start with 3, not 20: one for memory, one for tokens, one for your actual job run the same task before and after. keep only what makes the result better or cheaper every plugin you add loads into context, so more isn't free save this before you spend another weekend building something that already exists and repost it for the friend who's still paying $19 a skill ↓
56
191
11
1,525
163,722
Kalilur Rahman retweeted
the fastest growing GitHub repos in finance + AI this week: 1. Kronos (+2,653 ★) open source foundation model for candlestick charts, trained on data from 45+ exchanges. you feed it price history, it forecasts the next bars. code hasn't changed since april, people just found it. github.com/shiyu-coder/Krono… 2. TradingAgents (+1,944 ★) multi-agent LLM trading framework: analysts, researchers, a trader and a risk team arguing before every trade. #2 again, now 113K stars. github.com/TauricResearch/Tr… 3. day-trader-skill (+972 ★) 11 Claude skills for day trading. chart scanner, news tracker, entry and stop planner, and a "time machine" that replays the session 10,000 times before you trade. paper trading by default. 4 days old. github.com/Jakeschincariol/d… 4. Vibe-Trading (+821 ★) personal trading agent from HKUDS. research in plain english, strategy drafts, backtests, MCP support. github.com/HKUDS/Vibe-Tradin… 5. last30days-skill (+730 ★) agent skill that researches any topic across Reddit, X, YouTube, HN and Polymarket, and ranks results by upvotes, likes and real money. github.com/mvanhorn/last30da… 6. daily_stock_analysis (+487 ★) LLM stock analysis for US, China, HK and more. sends you a daily decision dashboard to Telegram, Discord or email. github.com/ZhuLinsen/daily_s… 7. FinceptTerminal (+412 ★) open source Bloomberg-style terminal. market data, research and AI agents in one desktop app. github.com/Fincept-Corporati… 8. freqtrade (+360 ★) the OG open source crypto trading bot, with a machine learning module (FreqAI). backtests, Telegram control, supports Hyperliquid. github.com/freqtrade/freqtra… 9. tick-stock-panel (+346 ★) self-hosted quant workbench for China A-shares. LLM builds your strategy, then backtest and monitoring. 65 API endpoints + an MCP server. github.com/shy3130/tick-stoc… 10. OpenBB (+294 ★) open data platform for analysts, quants and AI agents. stocks, options, crypto, macro. has an MCP server. github.com/openbq-org/OpenBB bookmark this and start today.
the fastest growing GitHub repos in finance + AI this week: 1. aurelio-finance (+2,388 ★) personal finance app with an AI advisor you can talk to. tracks net worth, ETFs, investments and cash. runs on your own computer, AI is optional with your own key. the repo is 11 days old. github.com/LosaLosSantos/aur… 2. TradingAgents (+1,244 ★) multi-agent LLM trading framework, analysts + researchers + trader + risk team in one stack. back on the list again. now 111K stars, new release lets it run on your ChatGPT Plus/Pro plan. github.com/TauricResearch/Tr… 3. last30days-skill (+735 ★) agent skill that researches any topic across Reddit, X, YouTube, HN and Polymarket, and ranks what it finds by upvotes, likes and real money. works in Claude Code, Codex and Cursor. github.com/mvanhorn/last30da… 4. Vibe-Trading (+621 ★) personal trading agent from HKUDS. market research in plain english, strategy drafts, backtests, MCP support. one command to give your agent trading skills. github.com/HKUDS/Vibe-Tradin… 5. nautilus_trader (+463 ★) Rust trading engine where the same strategy code runs in backtest and live. any asset, any venue with an API. biggest day of the week was friday, +130. github.com/nautechsystems/na… 6. daily_stock_analysis (+428 ★) LLM stock analysis for US, China, HK, Japan, Korea and Taiwan. sends you a daily decision dashboard to Telegram, Discord, Slack or email. github.com/ZhuLinsen/daily_s… 7. freqtrade (+342 ★) the OG open source crypto trading bot, now with a machine learning module (FreqAI). backtesting, Telegram control, supports Hyperliquid. github.com/freqtrade/freqtra… 8. OpenBB (+317 ★) open data platform for analysts, quants and AI agents. stocks, options, crypto, macro. one pip install, has an MCP server. github.com/openbq-org/OpenBB 9. FinceptTerminal (+293 ★) open source Bloomberg-style terminal. market analytics, research and AI agents in one desktop app. github.com/Fincept-Corporati… 10. QuantDinger (+266 ★) self-hosted AI trading OS. idea → python strategy → backtest → paper → live, with monitoring. was #6 back in may, still growing. github.com/OpenByteInc/Quant… bookmark this and start today.
145
263
25
2,543
327,103
Kalilur Rahman retweeted
Citadel and Renaissance make billions on one test you've never heard of. It's now free. And almost no trading strategy survives it Most traders don't even know their strategy is dead. Because they never do the one thing every hedge fund does Split the history in half. Build the strategy on the first half. Don't touch the second one, the strategy must never see it Then test it only on that second half. If the edge is gone, it was never there. It was a fit to the past But here's the trap few people know. If you tried 200 versions and kept the best one, even this check will lie to you. Because one of the 200 passes it by pure luck That's what funds count and you don't. Not how pretty the chart is, but how many times you tried before it got pretty I took a simple rule: hold the S&P when it's above its 200-day average, exit when it's below On the years it had never seen, the same growth as buy-and-hold, but half the drawdown Then I ran it with the average at 50, 100, 150, 200, 250 days. A real edge holds across all of them. Luck works on one number and dies on the next Most textbook strategies don't pass this test. Your meme-coin one has no chance And here's the main trap: the prettier the chart, the more you believe it. The more you believe, the more you bet So the strategy you love most takes the most money The market will test your strategy either way. The only question is whether it's before or after you put money into it How to run this test on real data, before your strategy touches your money, in the article below
13
19
151
12,073
Kalilur Rahman retweeted
🚨 ESTOS 10 REPOS DE TRADING ESTÁN CREANDO MILLONARIOS EN SILENCIO… Y SON 100% GRATIS > 10 repos open source > 435K+ estrellas en GitHub > $0 en licencias > 1 stack: pensar, probar, ejecutar 10 repos de trading open source que deberías tener guardados: AGENTES DE IA QUE ANALIZAN: 1. TradingAgents (111K ⭐) → github.com/TauricResearch/Tr… un equipo de agentes de IA: analistas, investigadores alcistas y bajistas, trader y gestor de riesgo que debaten cada operación 2. ai-hedge-fund (64K ⭐) → github.com/virattt/ai-hedge-… agentes que analizan acciones como lo harían Buffett, Munger o Burry. Educativo: no opera con dinero real BOTS QUE OPERAN: 3. freqtrade (55K ⭐) → github.com/freqtrade/freqtra… el bot de cripto gratis más usado: backtesting, optimización de estrategias y control desde Telegram 4. Hummingbot (20K ⭐) → github.com/hummingbot/hummin… bots de market making y arbitraje en exchanges de cripto MOTORES DE BACKTESTING: 5. NautilusTrader (30K ⭐) → github.com/nautechsystems/na… motor escrito en Rust: la misma estrategia corre en backtest y en vivo 6. Lean de QuantConnect (22K ⭐) → github.com/QuantConnect/Lean el motor de trading algorítmico de QuantConnect, en Python y C# IA PARA QUANTS: 7. Qlib de Microsoft (49K ⭐) → github.com/microsoft/qlib plataforma de inversión cuantitativa con IA, de los datos al modelo 8. FinRL (16K ⭐) → github.com/AI4Finance-Founda… aprendizaje por refuerzo aplicado a trading LA BASE: 9. ccxt (44K ⭐) → github.com/ccxt/ccxt una sola API para más de 100 exchanges de cripto y mercados de predicción 10. machine-learning-for-trading (21K ⭐) → github.com/stefan-jansen/mac… todo el código del libro de Stefan Jansen: de conseguir datos a operar en vivo Guárdalo antes de que sea tarde 🔖
FULL TUTORIAL TO BUILD YOUR OWN AI TRADING BOT WITH CLAUDE IN 14 MINUTES This ex-Meta engineer builds a bot that trades for you 24/7, and you don’t need to know how to code. → connects Claude Code to Alpaca with an MCP → gives it a simple strategy (20/50-day moving average crossover) → the bot checks the market on its own and buys or sells when the signal hits people sell courses like this for $497. he’s giving it away for free, and he tests everything with paper trading first. 🔖 Save it, it’s worth it.
39
226
7
1,398
185,768
Kalilur Rahman retweeted
this is pure f*cking gold Andrej Karpathy landed at Anthropic 20 weeks ago two things are already public, and together they're the whole playbook: > Anthropic's own knowledge graph cookbook > Karpathy's autoresearch loop: 700 experiments in 2 days, 20 optimizations found with nobody steering stack them and you get graph engineering on top of a loop i wired the graph approach into my own setup. the first reply already felt different Claude dropped the canned answer and reasoned through the problem step by step the loop runs the experiments. the graph decides what's worth remembering read the cookbook first, then the article below
10
46
286
16,217
Kalilur Rahman retweeted
this is pure f*cking treasure A Stanford AI research group found how to orchestrate Claude Opus 5.5 and GPT-6.1 Sol together, and it completely breaks the search scaffolding wall most developers try to combine frontier models by chaining prompts in one context window or hardcoding rigid evolutionary loops. you burn tokens on context bloat and freeze models into fixed search rules that out-date themselves within 5 iterations Stanford's architecture eliminates human-designed search scaffolding and splits the cognitive stack: > search director: Claude Opus 5.5 plans the search, querying a persistent idea-graph where MAP-Elites and MCTS reduce to single Cypher queries > candidate proposer: GPT-6.1 Sol explores high-entropy code variations inside isolated execution sandboxes > zero context bleeding: sessions reset after each iteration to prevent prompt bloat while the graph stores lineage > evaluator isolation: candidate code never touches the scoring process, stopping reward hacking and test leakage cold > meta-agent distillation: an offline meta-pass extracts verified lessons and updates search guidance across branches the benchmark metrics outclass standard discovery baselines: > 3.2x lower model spend than fixed evolutionary frameworks > 1st place rank across 7 competitive AtCoder heuristic contests against human competitors > beats published SOTA on Anthropic kernel builder and 11 mathematical optimization tasks > 0% reward hacks: candidate-controlled metric tampering eliminated Opus directs the search graph. Sol explores the code you stop writing rigid search harnesses. you let frontier models own the discovery loop
137
601
34
3,471
281,715
Kalilur Rahman retweeted
Jane Street just released how they use LLMs + AI agents for trading and research - I put together 16-page PDF the twist: they keeps even incorrect historical data - because a "corrected" history can create a trading edge that never existed here's the playbook in 6 steps: step 1 → pick one ticker, one hypothesis, one horizon - give your AI a specific question it can test step 2 → snapshot every source with its receipt time - corrections become new versions, never rewrite what you knew before a trade step 3 → make the LLM extract facts into JSON - every number carries its source, supporting passage, unit and reporting period step 4 → let code calculate and validate - wrong units, missing evidence or data arriving after the cutoff trigger BLOCK step 5 → test in time order against a simple baseline - include trading costs, keep later information out of earlier decisions step 6 → freeze predictions on new data before outcomes - log the misses, inspect failures, change one component and repeat the result: a chatbot becomes a trading research desk - facts have evidence, predictions have timestamps, and ideas face tests before risking money read this 16-page PDF with 10 prompts and the complete build workflow inside bookmark and build the full system from the article below ↓
36
100
2
570
65,321
Kalilur Rahman retweeted
OPUS 5.5 + DOTS + JEV closed 312 tickets in one week while 25 engineers closed 96... that org chart costs $343,750 a month in payroll the agent costs $2,055 a month in tokens, 167x less the code comes from Opus 5.5, the tests from Dots, and nothing reaches done without Jev's review only 410 of the team's 1,000 weekly hours are writing code, and that 41% is what the agent takes the other 590 are meetings, review, on-call, customers and deciding on the other side is a business billing $120,198 a month, 402 people paying $299 after fees, hosting, the agent and 3 humans, $66,887 is left before tax and $120,198 pays for 8.7 engineers at most, so that business never had 25 to replace ↓
15
23
146
19,408