The Curious Engineer

All blogs / AI / ML

AI Development Blog

augmentcode.com/blog

Product updates, insights, and research from Augment Code.

AI / MLSoftware / engineering practice

’24’25’26
Posts
114 from 26 Apr 2024 to 8 Oct 2026
Posting
weekly (58 in the last year)
Kind
company blog

2026 44

  1. Sep 11, 2026Beyond AI Coding Agents: How We Built Augment's Software FactoryAkshay UttureRead
  2. Oct 8, 2026Augment Code's next chapterMatt McClernanRead
  3. Karpathy skills on OpenClaw: agents don't write better code. But they do it more efficiently.Apr 30, 2026Karpathy skills on OpenClaw: agents don't write better code. But they do it more efficiently.Apr 30, 2026
  4. A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all.Apr 22, 2026A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all.Apr 22, 2026
  5. The bottleneck moved to verification. So we automated that too.Jul 1, 2026The bottleneck moved to verification. So we automated that too.Jul 1, 2026
  6. Agents can now run the full SDLC in Cosmos. What do engineers do? Jun 15, 2026Agents can now run the full SDLC in Cosmos. What do engineers do? Jun 15, 2026
  7. The era of single-model engineering is overApr 7, 2026The era of single-model engineering is overApr 7, 2026
  8. What is loop engineering, and how are leading software engineering teams using it?Jul 22, 2026What is loop engineering, and how are leading software engineering teams using it?Jul 22, 2026
  9. Introducing Augment Prism: model routing to reduce cost and maintain qualityMay 2, 2026Introducing Augment Prism: model routing to reduce cost and maintain qualityMay 2, 2026
  10. AI transformation is a systems problem: a conversation with Justin Reock, DXApr 28, 2026AI transformation is a systems problem: a conversation with Justin Reock, DXApr 28, 2026
  11. Excited, anxious, invigorated: what 219 engineering leaders told us about going AI-nativeMay 12, 2026Excited, anxious, invigorated: what 219 engineering leaders told us about going AI-nativeMay 12, 2026
  12. How Cosmos accelerated our AI-native transformation by removing the code review bottleneckMay 6, 2026How Cosmos accelerated our AI-native transformation by removing the code review bottleneckMay 6, 2026
  13. Scaling incident management for an AI-native organization using CosmosMay 26, 2026Scaling incident management for an AI-native organization using CosmosMay 26, 2026
  14. Opus 4.7 for 33% less: How Auggie beats Claude Code on cost and qualityMay 15, 2026Opus 4.7 for 33% less: How Auggie beats Claude Code on cost and qualityMay 15, 2026
  15. An AI analyst that reads Linear, Slack, and GitHub and answers team queries off our dbt modelsJun 29, 2026An AI analyst that reads Linear, Slack, and GitHub and answers team queries off our dbt modelsJun 29, 2026
  16. Accelerating large engineering projects with CosmosJun 18, 2026Accelerating large engineering projects with CosmosJun 18, 2026
  17. Using the Cosmos software factory to better understand our businessAug 13, 2026Using the Cosmos software factory to better understand our businessAug 13, 2026
  18. The best value for every token: GPT-5.6 Sol is now our default in CosmosJul 29, 2026The best value for every token: GPT-5.6 Sol is now our default in CosmosJul 29, 2026
  19. How we rebuilt the Auggie CLI harness to make tasks 53% cheaper than Claude CodeAug 14, 2026How we rebuilt the Auggie CLI harness to make tasks 53% cheaper than Claude CodeAug 14, 2026
  20. The software factory needs a faster review loop: further optimizing the path from PR to mergeAug 13, 2026The software factory needs a faster review loop: further optimizing the path from PR to mergeAug 13, 2026
  21. We think the FDE should be part of the product, so we built one inAug 28, 2026We think the FDE should be part of the product, so we built one inAug 28, 2026
  22. How a two-person engineering team handled 6× more product feedback and kept shippingAug 27, 2026How a two-person engineering team handled 6× more product feedback and kept shippingAug 27, 2026
  23. You don't have an AI problem. You have a transformation problem.Feb 28, 2026You don't have an AI problem. You have a transformation problem.Feb 28, 2026
  24. 1 spec isn't cool. You know what's cool? 10 specs.Feb 26, 20261 spec isn't cool. You know what's cool? 10 specs.Feb 26, 2026
  25. Context is the product. And now it’s portable.Mar 6, 2026Context is the product. And now it’s portable.Mar 6, 2026
  26. GPT-5.4 is now the default model in Augment and free for a limited time. Here’s why.Mar 5, 2026GPT-5.4 is now the default model in Augment and free for a limited time. Here’s why.Mar 5, 2026
  27. Why every AI coding breakthrough feels normal within 90 daysFeb 27, 2026Why every AI coding breakthrough feels normal within 90 daysFeb 27, 2026
  28. What spec-driven development gets wrongFeb 20, 2026What spec-driven development gets wrongFeb 20, 2026
  29. Frontier AI at half the cost: Gemini 3.1 Pro now available in Augment Code Apr 2, 2026Frontier AI at half the cost: Gemini 3.1 Pro now available in Augment Code Apr 2, 2026
  30. How we hire AI-native engineers now: our criteria Mar 12, 2026How we hire AI-native engineers now: our criteria Mar 12, 2026
  31. Your agent's context is a junk drawerFeb 27, 2026Your agent's context is a junk drawerFeb 27, 2026
  32. The hardest part about going AI-native isn't just technicalFeb 27, 2026The hardest part about going AI-native isn't just technicalFeb 27, 2026
  33. How I went AI-native in my terminal workflowMar 24, 2026How I went AI-native in my terminal workflowMar 24, 2026
  34. Review the spec, not the codeMar 16, 2026Review the spec, not the codeMar 16, 2026
  35. Specs are infrastructure in the age of agentsMar 13, 2026Specs are infrastructure in the age of agentsMar 13, 2026
  36. How we built a high-quality AI code review agentMar 10, 2026How we built a high-quality AI code review agentMar 10, 2026
  37. Building faster than you can thinkMar 9, 2026Building faster than you can thinkMar 9, 2026
  38. Prompts are infrastructure: building agents that actually listenMar 5, 2026Prompts are infrastructure: building agents that actually listenMar 5, 2026
  39. We rebuilt code review for how engineers want to work in 2026: infinite scroll and swipingApr 1, 2026We rebuilt code review for how engineers want to work in 2026: infinite scroll and swipingApr 1, 2026
  40. Confidence is the new bottleneckMar 27, 2026Confidence is the new bottleneckMar 27, 2026
  41. The AI-native transformation path, end to endApr 7, 2026The AI-native transformation path, end to endApr 7, 2026
  42. Designing with code, without writing codeApr 2, 2026Designing with code, without writing codeApr 2, 2026
  43. Code review at scale is broken. Here's how we're fixing it. Dec 11, 2025Code review at scale is broken. Here's how we're fixing it. Dec 11, 2025
  44. We benchmarked 7 AI code review tools on large open-source projects. Here are the results.Dec 11, 2025We benchmarked 7 AI code review tools on large open-source projects. Here are the results.Dec 11, 2025

2025 58

  1. Rebuilding state management: How we made our VS Code extension 2× fasterDec 18, 2025Rebuilding state management: How we made our VS Code extension 2× fasterDec 18, 2025
  2. Beyond adoption: How to measure AI's real business impactDec 18, 2025Beyond adoption: How to measure AI's real business impactDec 18, 2025
  3. Why GPT-5.2 is our model of choice for Augment Code ReviewDec 11, 2025Why GPT-5.2 is our model of choice for Augment Code ReviewDec 11, 2025
  4. Auggie supports ACP–now available in Zed, Neovim, and EmacsNov 6, 2025Auggie supports ACP–now available in Zed, Neovim, and EmacsNov 6, 2025
  5. Our new credit-based plans are now liveOct 21, 2025Our new credit-based plans are now liveOct 21, 2025
  6. Developers are choosing older AI models — and the data explain whyOct 20, 2025Developers are choosing older AI models — and the data explain whyOct 20, 2025
  7. Cultivating a culture of AI experimentation: A practical guide for engineering leadersJan 30, 2026Cultivating a culture of AI experimentation: A practical guide for engineering leadersJan 30, 2026
  8. The end of linear workJan 28, 2026The end of linear workJan 28, 2026
  9. Where to start with AI automation in the SDLC: practical advice for teams at scaleJan 23, 2026Where to start with AI automation in the SDLC: practical advice for teams at scaleJan 23, 2026
  10. Automate security at scale: How Augment Code and Snyk eliminate manual security bottlenecksJan 9, 2026Automate security at scale: How Augment Code and Snyk eliminate manual security bottlenecksJan 9, 2026
  11. Vibe Coding: Generating tech debt at the speed of lightJan 8, 2026Vibe Coding: Generating tech debt at the speed of lightJan 8, 2026
  12. Building by using: How each iteration of our CLI taught us what developers actually needDec 18, 2025Building by using: How each iteration of our CLI taught us what developers actually needDec 18, 2025
  13. Augment's Context Engine is now available for any AI coding agentFeb 6, 2026Augment's Context Engine is now available for any AI coding agentFeb 6, 2026
  14. Auggie tops SWE-Bench ProFeb 4, 2026Auggie tops SWE-Bench ProFeb 4, 2026
  15. We're in an exponentialFeb 12, 2026We're in an exponentialFeb 12, 2026
  16. Intent: A workspace for agent orchestrationFeb 10, 2026Intent: A workspace for agent orchestrationFeb 10, 2026
  17. Augment and Jellyfish partner to make AI impact measurableFeb 18, 2026Augment and Jellyfish partner to make AI impact measurableFeb 18, 2026
  18. Supercharge development in Augment with skills from Convex, Railway, Sentry, Supabase, and VercelFeb 13, 2026Supercharge development in Augment with skills from Convex, Railway, Sentry, Supabase, and VercelFeb 13, 2026
  19. Augment Code’s pricing is changing on October 20, 2025Oct 6, 2025Augment Code’s pricing is changing on October 20, 2025Oct 6, 2025
  20. Announcing New MCP Integrations: Vercel, Honeycomb, Render, and HerokuSep 16, 2025Announcing New MCP Integrations: Vercel, Honeycomb, Render, and HerokuSep 16, 2025
  21. How to build your agent: 11 prompting techniques for better AI agentsMay 21, 2025How to build your agent: 11 prompting techniques for better AI agentsMay 21, 2025
  22. Prompt Enhancer live in Augment ChatMay 14, 2025Prompt Enhancer live in Augment ChatMay 14, 2025
  23. Introducing Easy MCP: 1-click integration to CircleCI, MongoDB, Redis, Sentry, and Stripe Jul 30, 2025Introducing Easy MCP: 1-click integration to CircleCI, MongoDB, Redis, Sentry, and Stripe Jul 30, 2025
  24. Claude Sonnet 4: The best model. With the best Context Engine.May 22, 2025Claude Sonnet 4: The best model. With the best Context Engine.May 22, 2025
  25. Production-ready AI: Remote Agents now available for VS CodeJun 5, 2025Production-ready AI: Remote Agents now available for VS CodeJun 5, 2025
  26. Augment Code achieves ISO/IEC 42001 certification for its AI management systemMay 29, 2025Augment Code achieves ISO/IEC 42001 certification for its AI management systemMay 29, 2025
  27. Rolling out AI coding assistants: How Drata did itJul 10, 2025Rolling out AI coding assistants: How Drata did itJul 10, 2025
  28. How we made code search 40% faster for 100M+ line codebases using quantized vector searchJun 11, 2025How we made code search 40% faster for 100M+ line codebases using quantized vector searchJun 11, 2025
  29. From spaghetti prompts to structured plans: Tasklist in real-world engineeringJul 28, 2025From spaghetti prompts to structured plans: Tasklist in real-world engineeringJul 28, 2025
  30. How we solved the "AI agent black box" problem with typed tasksJul 28, 2025How we solved the "AI agent black box" problem with typed tasksJul 28, 2025
  31. Why the future of software engineering is beyond the IDEJul 31, 2025Why the future of software engineering is beyond the IDEJul 31, 2025
  32. Context Engine. Now with full Commit historyJul 29, 2025Context Engine. Now with full Commit historyJul 29, 2025
  33. GPT-5 is here. And we now have a Model PickerAug 7, 2025GPT-5 is here. And we now have a Model PickerAug 7, 2025
  34. Context beats modeling: 45% less code typing due to intent-based completions modelAug 6, 2025Context beats modeling: 45% less code typing due to intent-based completions modelAug 6, 2025
  35. Using the Auggie CLI for automated code reviewSep 4, 2025Using the Auggie CLI for automated code reviewSep 4, 2025
  36. Automating customer feedback and support with AI AgentsSep 3, 2025Automating customer feedback and support with AI AgentsSep 3, 2025
  37. Augment is now more affordable. Introducing our $20 per month Indie planSep 12, 2025Augment is now more affordable. Introducing our $20 per month Indie planSep 12, 2025
  38. How we built Memory ReviewSep 8, 2025How we built Memory ReviewSep 8, 2025
  39. New, simpler pricing: Pay for what you control. Enjoy everything else.May 7, 2025New, simpler pricing: Pay for what you control. Enjoy everything else.May 7, 2025
  40. To fork or not to fork?Mar 7, 2025To fork or not to fork?Mar 7, 2025
  41. The rise of specialty models: 6 predictions for AI in 2025Feb 7, 2025The rise of specialty models: 6 predictions for AI in 2025Feb 7, 2025
  42. Securing the code that writes code: A look inside Augment Code's AI platformFeb 3, 2025Securing the code that writes code: A look inside Augment Code's AI platformFeb 3, 2025
  43. Introducing Next Edit: AI that understands the ripple effect of code changesFeb 19, 2025Introducing Next Edit: AI that understands the ripple effect of code changesFeb 19, 2025
  44. Augment <3 VimFeb 14, 2025Augment <3 VimFeb 14, 2025
  45. AI model pickers are a design failure, not a featureMar 6, 2025AI model pickers are a design failure, not a featureMar 6, 2025
  46. The AI research behind Next EditFeb 19, 2025The AI research behind Next EditFeb 19, 2025
  47. Image support comes to Augment Chat for VS CodeMar 19, 2025Image support comes to Augment Chat for VS CodeMar 19, 2025
  48. No forks allowed: How user-centric design saved Next Edit from compromiseMar 18, 2025No forks allowed: How user-centric design saved Next Edit from compromiseMar 18, 2025
  49. Best practices for using AI coding AgentsApr 2, 2025Best practices for using AI coding AgentsApr 2, 2025
  50. Meet Augment Agent: Your AI pair programmer that deeply understands your codebase, and learns as you workApr 2, 2025Meet Augment Agent: Your AI pair programmer that deeply understands your codebase, and learns as you workApr 2, 2025
  51. #1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1Mar 31, 2025#1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1Mar 31, 2025
  52. You make your evals, then your evals make you. Introducing AugmentQAMar 26, 2025You make your evals, then your evals make you. Introducing AugmentQAMar 26, 2025
  53. Customer Managed Keys: Your keys, your rulesApr 16, 2025Customer Managed Keys: Your keys, your rulesApr 16, 2025
  54. How to prompt your Agent: 8 tips on improving AI agent performanceApr 9, 2025How to prompt your Agent: 8 tips on improving AI agent performanceApr 9, 2025
  55. Software agents you can trustMay 2, 2025Software agents you can trustMay 2, 2025
  56. Augment leads on CCEval: Benchmarking code completion for continuous improvementApr 20, 2025Augment leads on CCEval: Benchmarking code completion for continuous improvementApr 20, 2025
  57. Remote Agent clears your backlog while you plan what’s nextMay 7, 2025Remote Agent clears your backlog while you plan what’s nextMay 7, 2025
  58. 4 Essential data questions for your AI coding platformMay 6, 20254 Essential data questions for your AI coding platformMay 6, 2025

2024 12

  1. Augment Code raises $227 Million to empower software teams with AIApr 26, 2024Augment Code raises $227 Million to empower software teams with AIApr 26, 2024
  2. Beyond code generation: What if your AI actually understood your codebase?Dec 18, 2024Beyond code generation: What if your AI actually understood your codebase?Dec 18, 2024
  3. Personal journey: Reimagining software engineering through AIAug 15, 2024Personal journey: Reimagining software engineering through AIAug 15, 2024
  4. Introducing Augment: a company dedicated to empowering developers with AIAug 1, 2024Introducing Augment: a company dedicated to empowering developers with AIAug 1, 2024
  5. Webflow developers stay in the flow with Augment’s context-aware AI suggestionsOct 15, 2024Webflow developers stay in the flow with Augment’s context-aware AI suggestionsOct 15, 2024
  6. Augment Code achieves SOC2 Type IISep 5, 2024Augment Code achieves SOC2 Type IISep 5, 2024
  7. Rethinking LLM inference: Why developer AI needs a different approachNov 14, 2024Rethinking LLM inference: Why developer AI needs a different approachNov 14, 2024
  8. Meet Augment Code: Developer AI for teamsOct 24, 2024Meet Augment Code: Developer AI for teamsOct 24, 2024
  9. A real-time index for your codebase: Secure, personal, scalableJan 1, 2025A real-time index for your codebase: Secure, personal, scalableJan 1, 2025
  10. Reinforcement learning from developer behaviors: A breakthrough in code generation qualityNov 26, 2024Reinforcement learning from developer behaviors: A breakthrough in code generation qualityNov 26, 2024
  11. The hidden cost of code complexity: why developer ramp-up takes longer than you thinkJan 23, 2025The hidden cost of code complexity: why developer ramp-up takes longer than you thinkJan 23, 2025
  12. How coding AIs will support large-scale software engineeringJan 7, 2025How coding AIs will support large-scale software engineeringJan 7, 2025