Startups On Rails 2025 – Sin City Ruby

Transcript

1. Irina Nazarova Startups on Rails 2025 1
3. 2 ON THE NEWS
4. 3 March 16, 2025
5. 3 March 16, 2025
6. RubyLLM on HN front page 3 March 16, 2025
7. RubyLLM on HN front page Carmine Paolino Author, founder Chatwithwork Berlin 3 March 16, 2025
8. 4 March 26, 2025
9. 4 March 26, 2025
10. 4 March 26, 2025
11. 4 March 26, 2025
12. 4 March 26, 2025
13. 4 March 26, 2025
14. 4 YC partner pitches Rails March 26, 2025
15. 5 January 22, 2025
16. .new Stack: Remix, Wasm, Rails 5 January 22, 2025
17. .new Stack: Remix, Wasm, Rails 5 January 22, 2025
18. .new Stack: Remix, Wasm, Rails Albert Pai and Eric Simmons Co-founders San Francisco 5 January 22, 2025
19. 6 February 13, 2025
20. 6 February 13, 2025
21. 6 February 13, 2025 raised
22. $1M? 6 February 13, 2025 raised
23. $1M? 6 5M? February 13, 2025 raised
24. $1M? 6 5M? 50M 🤯 February 13, 2025 raised
25. Nikita Savrov co-founder&CTO Maryland, US $1M? 6 5M? 50M 🤯 February 13, 2025 raised
26. Nikita Savrov co-founder&CTO Maryland, US $1M? 6 -Ruby on Rails monolith with Turbo -Experimental mobile app on Turbolinks -JS frameworks, microfrontends and Flutter in the mix 5M? 50M 🤯 February 13, 2025 raised
27. 7 March 3, 2025
28. 7 March 3, 2025
29. 7 plans an IPO March 3, 2025
30. 7 Ryan King Co-founder&CTO San Francisco plans an IPO March 3, 2025
31. 7 Ryan King Co-founder&CTO San Francisco plans an IPO March 3, 2025
32. 7 Ryan King Co-founder&CTO San Francisco plans an IPO March 3, 2025
33. 7 Ryan King Co-founder&CTO San Francisco plans an IPO March 3, 2025
35. 8 SO, WHERE ARE WE?
36. 8 SO, WHERE ARE WE? Startups on Rails are growing
37. 8 SO, WHERE ARE WE? Startups on Rails are growing Ruby open source is getting popular
38. 8 SO, WHERE ARE WE? Startups on Rails are growing Ruby open source is getting popular Investors support Rails
39. 9 Visibility of Technology Gartner Hype Cycle Time
40. 9 Visibility of Technology Gartner Hype Cycle Time
41. 9 Peak of in fl ated expectations Visibility of Technology Gartner Hype Cycle Time
42. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment Gartner Hype Cycle Time
43. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment S lop e of enlightenm ent Gartner Hype Cycle Time
44. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Gartner Hype Cycle Time
45. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Gartner Hype Cycle Time
46. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Pragmatic adoption Gartner Hype Cycle Time
47. 9 Peak of in fl ated expectations Visibility of Technology Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Pragmatic adoption Gartner Hype Cycle Time
48. Time for intros! Irina (Ira) Nazarova 10
49. Time for intros! Irina (Ira) Nazarova 10 CEO at Evil Martians
50. Time for intros! Irina (Ira) Nazarova 10 CEO at Evil Martians Co-founder at AnyCable
51. Time for intros! Irina (Ira) Nazarova 10 CEO at Evil Martians Co-founder at AnyCable Org of SF Ruby meetup
53. 11 LET’S HEAR FROM STARTUP FOUNDERS
57. 13 NOT THESE FOUNDERS
59. 14 NOT THESE FOUNDERS
108. 18 WHY RUBY AND RAILS?
109. 19 “SOMETHING I SPENT HALF A YEAR BUILDING WITH JAVA, I BUILT IN ONLY 2 MONTHS… WHEN LEARNING RAILS” Thiago Diniz Speed to Market
110. 20 “AFTER TRYING SCALA AND OTHER FRAMEWORKS, RAILS ALLOWED ME TO FOCUS ON BUSINESS LOGIC RATHER THAN CONFIGURATION” Bjarke Vad Andersen Convention over configuration
111. 21 “RAILS IS THE GROWN - UP IN THE ROOM: IT IS BATTLE - TESTED WITH YEARS AND YEARS OF CARE PUT INTO IT” Trent Kocurek Battle-Tested Reliability
112. 22 “RAILS DEVS TEND TO TAKE ON PM TASKS AND UNDERSTAND THE BUSINESS NEEDS” Aaron Snyder Entrepreneurial culture
113. 23 "I LIKE THE MAJESTIC MONOLITH" Ian Se ff erman Monolith Sanity
114. 24 “HOTWIRE SOLIDIFIED RAILS FOR THE NEXT 10 YEARS” Ryan Kulp Modern Frontend Solutions
115. 25 “RUBY IS OVER - REPRESENTED IN THE STARTUPS THAT GOT SUCCESSFUL” Kelly Sutton Startup Success Correlation
116. 26 “A HUGE UNLOCK OF CREATIVITY IN USING AI IS GOING TO HAPPEN IN FLEXIBLE LANGUAGES LIKE RUBY” Scott Werner Vision in AI
117. 27 “EVENTUALLY PEOPLE WILL REALIZE THAT THESE SUCCESSFUL COMPANIES HAVE BEEN BUILDING ON RAILS ALL ALONG” Glenn Espinosa Rails is a secret superpower
119. 28 WHAT’S MISSING?
120. 29 Peak of in fl ated expectations Visibility Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Pragmatic adoption Gartner Hype Cycle Time
121. 30 LET’S DO PRAGMATIC ENLIGHTENMENT Post-hype Rails We don’t need to be purists or over-hype anything
122. 31 IT’S UP TO US Post-hype Rails We can build the missing pieces and integrate them into Rails while preserving the DX and productivity
123. 32 Post-hype Rails
124. 32 Post-hype Rails
126. 33 TOP-4 MISSING
127. 34 #4 Frontend Experience
128. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience
129. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”)
130. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”) -High-quality UI components (“ShadCN for Rails”)
131. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”) -High-quality UI components (“ShadCN for Rails”) -Patterns for building reactive interfaces
132. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”) -High-quality UI components (“ShadCN for Rails”) -Patterns for building reactive interfaces -Design aesthetic of sca ff olds and UI kits
133. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”) -High-quality UI components (“ShadCN for Rails”) -Patterns for building reactive interfaces -Design aesthetic of sca ff olds and UI kits -Integration with JS frameworks (Inertia)
134. 34 Polished, documented, and cohesive frontend story: #4 Frontend Experience -Better Hotwire documentation and best practice (“it takes weeks until it clicks”) -High-quality UI components (“ShadCN for Rails”) -Patterns for building reactive interfaces -Design aesthetic of sca ff olds and UI kits -Integration with JS frameworks (Inertia) ⚡ hotwire.io ⚡ rubyui.com ⚡ inertia-rails.dev
135. 35 #3 Performance And Scale
136. 35 Prescriptive guidance or tooling: #3 Performance And Scale
137. 35 Prescriptive guidance or tooling: #3 Performance And Scale -Can we have a way to achieve under 100ms response time for any Rails app?
138. 35 Prescriptive guidance or tooling: #3 Performance And Scale -Can we have a way to achieve under 100ms response time for any Rails app? -Parallel processing is important for heavily I/O-bound chats with LLMs
139. 35 Prescriptive guidance or tooling: #3 Performance And Scale -Can we have a way to achieve under 100ms response time for any Rails app? -Parallel processing is important for heavily I/O-bound chats with LLMs ⚡ github.com/socketry/async ⚡ byroot.github.io ⚡ github.com/socketry/ async-cable ⚡ github.com/anycable/ anycable
140. 36 #2 AI Tooling
141. 36 #2 AI Tooling THIS IS STILL TOO EARLY
142. 36 #2 AI Tooling THIS IS STILL TOO EARLY NO, IT’S TOO LATE!
143. 36 #2 AI Tooling THIS IS STILL TOO EARLY NO, IT’S TOO LATE! ⚡ github.com/crmne/ ruby_llm ⚡ github.com/sublayerapp/ sublayer ⚡ github.com/patterns-ai- core/langchainrb
144. 37 #2 AI Tooling
145. 37 Integration with “AI middleware” #2 AI Tooling
146. 37 Integration with “AI middleware” #2 AI Tooling -Connectivity between Rails monolith and Python services built by data engineers “AI middleware”
147. 37 Integration with “AI middleware” #2 AI Tooling -Connectivity between Rails monolith and Python services built by data engineers “AI middleware” Data engineering & Data Science Product engineering
148. 38 #2 AI Tooling
149. 38 LLM Evals in Ruby #2 AI Tooling
150. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning #2 AI Tooling
151. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning -Internal dashboard to monitor performance #2 AI Tooling
152. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning -Internal dashboard to monitor performance -Upload a dataset to check against locally #2 AI Tooling
153. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning -Internal dashboard to monitor performance -Upload a dataset to check against locally -Run it on production to check against production data #2 AI Tooling
154. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning -Internal dashboard to monitor performance -Upload a dataset to check against locally -Run it on production to check against production data #2 AI Tooling
155. 38 LLM Evals in Ruby -Gem to test prompts, with built-in prompt versioning -Internal dashboard to monitor performance -Upload a dataset to check against locally -Run it on production to check against production data #2 AI Tooling ⚡ github.com/kieranklaassen/leva
156. 39 RAG tooling #2 AI Tooling
157. 39 RAG tooling -Generating and managing embeddings #2 AI Tooling
158. 39 RAG tooling -Generating and managing embeddings -Rails-native tools for document ingestion, chunking, and preprocessing #2 AI Tooling
159. 39 RAG tooling -Generating and managing embeddings -Rails-native tools for document ingestion, chunking, and preprocessing -Active Record embedding? #2 AI Tooling
160. 39 RAG tooling -Generating and managing embeddings -Rails-native tools for document ingestion, chunking, and preprocessing -Active Record embedding? # A concept of Active Record Embedding class Post < ApplicationRecord has_one_attached :pdf has_embedding :content, service: :mini_llm has_embedding :pdf, service: :openai, chunked: {recursive: true, engine: :baran} end # Usage example post.content_embedding.vector # Returns vector post.pdf_chunks # Return chunks w/ vectors Post.with_similar_content(query, limit: 5) # Search similar records #2 AI Tooling
161. 39 RAG tooling -Generating and managing embeddings -Rails-native tools for document ingestion, chunking, and preprocessing -Active Record embedding? # A concept of Active Record Embedding class Post < ApplicationRecord has_one_attached :pdf has_embedding :content, service: :mini_llm has_embedding :pdf, service: :openai, chunked: {recursive: true, engine: :baran} end # Usage example post.content_embedding.vector # Returns vector post.pdf_chunks # Return chunks w/ vectors Post.with_similar_content(query, limit: 5) # Search similar records #2 AI Tooling ⚡ github.com/patterns-ai-core/langchainrb
162. 39 RAG tooling -Generating and managing embeddings -Rails-native tools for document ingestion, chunking, and preprocessing -Active Record embedding? # A concept of Active Record Embedding class Post < ApplicationRecord has_one_attached :pdf has_embedding :content, service: :mini_llm has_embedding :pdf, service: :openai, chunked: {recursive: true, engine: :baran} end # Usage example post.content_embedding.vector # Returns vector post.pdf_chunks # Return chunks w/ vectors Post.with_similar_content(query, limit: 5) # Search similar records #2 AI Tooling ⚡ github.com/patterns-ai-core/langchainrb ⚡ github.com/ankane/neighbor
163. 40 #2 AI Tooling
164. 40 AI agent tooling #2 AI Tooling
165. 40 AI agent tooling -Work fl ow orchestration: long running, durable, resumable, testable, performant work fl ows #2 AI Tooling
166. 40 AI agent tooling -Work fl ow orchestration: long running, durable, resumable, testable, performant work fl ows -Agent memory and conversational state management #2 AI Tooling
167. 40 AI agent tooling -Work fl ow orchestration: long running, durable, resumable, testable, performant work fl ows -Agent memory and conversational state management -Tracking AI agent performance, logging interactions, and analyzing response quality #2 AI Tooling
168. 40 AI agent tooling -Work fl ow orchestration: long running, durable, resumable, testable, performant work fl ows -Agent memory and conversational state management -Tracking AI agent performance, logging interactions, and analyzing response quality ⚡ github.com/ fractaledmind/ acidic_job #2 AI Tooling
169. 40 AI agent tooling -Work fl ow orchestration: long running, durable, resumable, testable, performant work fl ows -Agent memory and conversational state management -Tracking AI agent performance, logging interactions, and analyzing response quality ⚡ github.com/ fractaledmind/ acidic_job ⚡ activeagents.ai #2 AI Tooling
170. 41 #2 AI Tooling
171. 41 MCP tooling #2 AI Tooling
172. 41 MCP tooling MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app #2 AI Tooling
173. 41 MCP tooling How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app #2 AI Tooling
174. 41 MCP tooling How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app Stripe, GitHub, Gitlab, Block! #2 AI Tooling
175. 41 MCP tooling -Ruby SDK to build an MCP server How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app Stripe, GitHub, Gitlab, Block! #2 AI Tooling
176. 41 MCP tooling -Ruby SDK to build an MCP server -MCP generator from Swagger How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app Stripe, GitHub, Gitlab, Block! #2 AI Tooling
177. 41 MCP tooling -Ruby SDK to build an MCP server -MCP generator from Swagger -MCP proxy How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app Stripe, GitHub, Gitlab, Block! #2 AI Tooling
178. 41 MCP tooling -Ruby SDK to build an MCP server -MCP generator from Swagger -MCP proxy How many of the first MCPs are built for Ruby backends? MCP is a “self-documented API” protocol that an LLM-operated system can use to interact with your app Stripe, GitHub, Gitlab, Block! ⚡ github.com/yjacquin/fast-mcp ⚡ github.com/funwarioisii/mcp-rb #2 AI Tooling
179. 42 “WE NEED A NEW REASON TO BE VOCAL ABOUT USING RUBY – DESPITE THE SCEPTICISM” #1 …
180. 43 Peak of in fl ated expectations Time Visibility Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Gartner Hype Cycle #1 …
181. 43 Peak of in fl ated expectations Time Visibility Trough of disillusionment S lop e of enlightenm ent Plateau of productivity Gartner Hype Cycle REMEMBER: SLOPE OF ENLIGHTENMENT 💡🕯☀ #1 …
182. 44 #1 Pragmatic Enlightenment
183. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment
184. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials
185. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials -We need more Ruby on Rails bloggers on YouTube
186. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials -We need more Ruby on Rails bloggers on YouTube
187. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials -We need more Ruby on Rails bloggers on YouTube ⚡ GoRails, Ken Greeff, SupeRails on YouTube
188. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials -We need more Ruby on Rails bloggers on YouTube ⚡ GoRails, Ken Greeff, SupeRails on YouTube ⚡ “Patterns of application development using AI” Obie Fernandez
189. 44 Let’s share success stories that demonstrate enduring value of Rails #1 Pragmatic Enlightenment -Companies: talk about Ruby and Rails in your technical blogs and mention it in press releases and on socials -We need more Ruby on Rails bloggers on YouTube ⚡ GoRails, Ken Greeff, SupeRails on YouTube ⚡ “Patterns of application development using AI” Obie Fernandez
190. 45 Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
191. 45 Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
192. 45 Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
193. 45 Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
194. 45 Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
195. 45 ⚡ rubyconferences.org ⚡ rubyvideo.dev Last 12 months Ruby conferences: 38 in 34 locations Ruby meetups: 639 in 127 locations #1 Pragmatic Enlightenment
198. 46 LET’S ADD SOME NEW NAMES
199. 47 #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
200. 47 WHAT ARE YOU BUILDING NEXT? #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
201. 47 WHAT ARE YOU BUILDING NEXT? #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
202. 47 WHAT ARE YOU BUILDING NEXT? #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
203. 47 WHAT ARE YOU BUILDING NEXT? #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
204. 47 WHAT ARE YOU BUILDING NEXT? #3 Performance And Scale #1 Pragmatic Enlightenment #4 Frontend Experience #2 AI Tooling
205. Thank you! Reach out: inazarova@evilmartians.com inazarova on X, Bluesky Irina Nazarova on Linkedin 🥷 Work with evilmartians.com ⚡Use anycable.io ❤Join lu.ma/sfruby 48