I'm an engineering student, programmer, 3d-artist, illustrator, musician and aspiring entreprenour from Finland. I have always dreamed of working at an innovative Silicon Valley startup. I'm facinated by tech, startups, math, physics, science, sowtware, lean, psychology and history.
.png)
I started programming at age 14. I can write programming and scripting languages like: python, html, css, javascript, java, xml, c#, Visual Basic and many others. I can also do some PLC-programming particulary using languages like ST and FBD using an environment like Codesys. I also have little bit of experience with node.js and opc ua.
I made a sheet based music app for android called BoxSound420. This app contains around 6000 lines of html, css and javascript and hundreds of lines of java and xml. I made the app in android studio but obviously Apache Cordova and React Native etc. could have also been used.
I also made a mobile app called FastRead that enables people to read PDF-books extremely fast because the words are presented individually on the screen, thus eliminating excess time spent on eye movement. This app was made using java and Android Studio.
I'm also working on couple games in the unity engine and my own new type of social media app. Stay tuned for more!
.png)
I have extenssive experience using blender. I can create highly realistic textured objects and scenes that can be used in games, art, movies etc. I mostly get my textures from textures.com
I can also do complex modelling and simulations in cad programs like Solid Edge & keyshot, Creo Parametric, AutoCAD, etc. I can model parts, assemblies and design engineering blueprints.
I have always been really good at understanding and drawing 3D-shapes and I taught people how to draw 3D-cars while I was at third grade. I can do realistic illustrations using paper or professional art programs like ibisPaint, Photoshop, Gimp and others.
I have made a lot of my own music using Waveform 11 wich is a professional level music software. My favorite VST-plugins are Synth 1, Surge and Vital. I have also used tools like Caustic, Bosca Ceoili and my own music app BoxSound420.
In the summer of 2022 I had an internship at Valmet flow laboratory where I conducted testing and reporting on the performance of valves and pneumatic actuators.
During my work I wrote a test program that reduced the theoretical testing time of one endurance test by 150 days.
I also made significant Visual Basic improvements to the workload input and workload analysis Excel-files used at Helsinki and Shanghai laboratories.
During my work I also developed an Excel VBA program for analyzing the average quarterly and annual project times, wich will help Valmet in their goal of reducing their average valve R&D project times by up to 20%.
I also spent two summers working at the Neles machine shop. During my work I mostly operated a Mazak lathe with mazatrol NC-control and Frömag broaching machine. I machined tough materials like acid resistant steel and spray coatings like NiBo (Nickel Boron) I also briefly used a Cazeneuve branded manual lathe for basic workpiece preparation work. The smallest tolerances were as low as 2μm
In the future I want to work in the US tech industry in Silicon Valley. I want to run an innovative silicon valley startup that aims to change the world and I want to build useful products and services for as many people as possible. Indeed, I consider helping people to be the most satisfying thing in the world. Because we are here only for so long it's the legacy that we leave behind that matters.
I'm majoring in mechanical- and minoring in electrical engineering at Aalto University in Espoo Finland. In the future I plan to learn more computer science, accounting and leadership.
I believe reading books is important and the best books are the ones we use for self developmet. I have read many books about Silicon Valley founders and believe building a strong mindset for the future is incredibly important - it ennables us to create an exciting new chapter for humanity. This Is what I call the theory of controlled influence.
Edit: This website was made years ago & hasn't been updated in a long time as it's much more important to build product than have a fancy website. I will proceed to rant below all the brilliant business ideas Yc missed in case for some reason they didn't like our main idea that we have demoed to Saab and innitially agreed to deploy one day in frontlines with the help of large foundation that operates in Ukraine. (NONE of these ideas were included in our application as we have been laser focused on building the current main product.)
A new type of fully graphical, automatically adjusting interface AI search engine to finally replace Google's blue links once and for all! I tried out some AI text-based search engines. The most popular such tool yielded such results: I asked: give me tech AI notifications at 6pm, couldn't do it; give me previews of the Spotify pages of 10 most popular playlists, can't do it; give image search on cats, images behind separate clicks, can't do it; change background to light green, can't do it; make me aninmation, can't do it. Many of these are easy to build&implement! While some may have users, I think these are far from any Google replacement. The AI search engine shouldn't just produce a wall of text with citations. While "your search engine result is a boring research paper" may appeal to the creators of such tools that come from research backgrounds, this won't fully appeal to the general public. For an example: What if the tool, through multiple iterations with users, learned that for search: "what year Eiffel Tower created" the optimal preferred answer would be to answer the question in the most optimally descriptive short phase and then provide the most useful contexts like show pictures from different construction phases in a timeline-like manner, thus yielding maximal user satisfaction. What is needed is a search engine that can do everything the user needs and, based on user behavior and deep psychological understanding of user, produce the ideal paragraphs, fonts, shapes; sleek interface with graphs, images, timelines, animations etc. tailored to user preference, all of which other tools place behind many other searches and multiple slow loading page-to-page clicks. This is what's needed to fully replace Google in the search engine business. (Provide user with few both bad and good results early on to establish baseline behaviors; proceed to analyze user satisfacttion to results that are produced with some randomization; apply "liking this" assumption to new users in smilar peer groups and re-run the algorithm!)
Peer-to-peer package service & online bidding. Companies like Airbnb and Uber show us that much better efficiency can be achieved in peer-to-peer services. Japan did an intersting experiment: using slime mold to plan the most optimal paths of Tokyo metro. Living creatures helped to find the most optimal paths!! And what better problem to solve than global logistics? Massive container ships & shipping centers result in extremely suboptimal path where packages travel back and forth. Online shoppers often have to wait massive time for shipments this is especiually bad with chinese sites like aliexpress. The argument is that a random person that just happens to have the same work trip as the package track will be significantly more efficient, cheaper, and faster than a package moving between multiple massive container ships trucks and holding centers back and forth as they have to track the path of millions of totally unrelated packages. Furthermore, people should be allowed to offer a bidding service. Let’s say Josh has a truck and can connect with users that pool together and bid for shipping. Let’s say people have long shipping times and pool together a better, cheaper, and faster option in a bidding platform. Basically, a crowd-sourced shipping option for every e-commerce website. This could be a huge help to solve excess capacity and speed! Let’s say you have a truck going from A to B; you can offer the extra space for random users online, thus radically improving logistics and the global efficiency of society.
Short-form video platform where user preference is determined from multiple (4) sliding axes. For example: If a user likes to see more humorous content, they slide down - if more serious, slide up. Want to see more human/animal content? They slide bottom left - more technology content, they slide top right. More lesser-known creators slide left - big creators slide right. The user will even have the ability to swap out axes, so they can have the ideal 4 axes of preference. Then, using this, we can use AI to ultra deeply undertand user psychology, ultra deeply understand user patterns and likes, so we know to show this and this type of content at this time of day etc. Much happier users. Functionality gives much better potential for ad targeting than any other shortform video platform like YouTube or Tiktok. The app would be an ultra-important TikTok replacement to protect US interests & counter harmful Chinese government influence on US youth. (a feel good story?)
Directory metadata context RAM for computers. Basically, a new piece of computer hardware that, through AI, has an intelligent understanding of file structures and how users interact with them, and through intelligent AI and algorithms stores key parts of metadata in a fast-to-access way to speed up all file access. Furthermore, it could be a key part in improving the efficiency of large LLM systems' training and usage.
A Linux AI cyberdeck-style mini computer the size of a Nintendo DS. Much smaller finger movement means significantly faster coding (good snappy feel), especially when combined with AI (also red rubber nub mouse like in old thinkpads). Super snappy, fast custom Linux interfaces and support so can comfortably play light Steam games on the go. The ultimate smartphone/lightweight laptop replacement for AI-assisted programming and lightweight gaming and on-the-go!
Copyright © 2022- Elias Puolakka
home 
