The Art Of Casually Building Amanvir Com

Emily Johnson
-
the art of casually building amanvir com

I really enjoyed speaking about "casually building" at Startup Shell, my university's flagship startup incubator. If you're someone who's interested in working on your own projects or products, I'd encourage you to watch the recording of the talk: If you'd prefer to just read through the slides instead, you can do that as well: I've tried my best to pack a lot of value into this talk: I include several examples from my own journey, and the talk's filled with advice I'd love to have received when I... If you liked this post, feel free to check out azigy, the app I'm currently working on. You can also take a look at "Taming LLMs with Structured Outputs", the talk I gave a week after this one at DevIgnition 2024, a Washington D.C.-based developer conference!

My name is Amanvir Parhar, but you can call me Aman. I like to code, write, and build things. If you'd like to reach out to me, please send me an email (amanvir.com [at] gmail [dot] com), or find me X. A talk I gave at DevIgnition 2024, a Washington D.C.-based developer conference. A talk I gave at Startup Shell, the University of Maryland's flagship startup incubator. At DevIgnition 2024, I gave a talk on using structured outputs to get better responses from large language models (LLMs).

I was particularly enthusiastic about this topic, as I've been working with LLMs for a while now, and I've found that structured outputs can be a game-changer when integrating LLMs into your software. You can watch a recording of the talk here: If you'd prefer to just skim through the slides instead, you can do that as well: This was a big personal accomplishment for me, as this was my first time giving a technical presentation of this kind, especially on this scale. LLMs are bad at picking random numbers. I invented a whole new sampling technique to fix this.

Using Richard Dawkins' famous evolutionary simulation to create a text scramble effect. A fun project that I learned a lot from. How I used Google's Gemini to fix typesetting and formatting issues in Project Gutenberg eBooks. My journey implementing a new error-diffusion dithering algorithm in Racket. ��c$^��Z��k��!�ؘ$%��b\{/�tS�G�*��'w��X��B���=�@���6c\�Ӂ�3U�4w =�*O��:K�%N BY�Ϋ�q�Lj��c����g��I�c���$�тf��L�j>����c��v�{��Zn����a�w�/ ��F��B>���-|�����Rc ��u��h�o]�J++�T�o:}���zų)���r�Z�J0��U����5<{ ƴ����e�hj.��P��/Dž���xA�9iF�g8��H$��{��� ���a �(��o�ʹ7H[�J�\�r�9�H�&�.Ǟ�� Š3����6b�)2��j�&���V5����̨t!�m��Ȥ}K�Xjt��Y�z�[�-�t�[CQ�����ҽG���c%��̨��p-Nek����|�&�,v"�P��<͏�dS_]��<����լ�@��*��瓕8�Vp�q���|� �H��$�� C�B�yB�FFqײ�b���>�rC#B{4�-�zKh ���;�~8kt�ω��7i��үr��fS�:�9�� h��՝1���Ͱ�Y�)�% m(@�0p��d���Q�j�f�גP?k�_�Ϯ%��}���`f��K�%�l*<.���20��?Y C��p$xQ�� �J����RZ*T�G/>E�@��}�>`��+����|\��z��Db���a��)k�$2�@!i0�-](�*8���K�� ���I7W�=��mm~b5kH�& )�C�Igu���}�) �&hoh6K���?/��z�ƒ6�ܕ|'�ށb�;ϣ�B�V&-�$�]嶛Oj S�'EևyP�[W���M7��4��\{�nu���߹�2^N�sꙞi�����r1K1��`��a����a�2��˺�2?�w�����Q(�_��f���S�P�_��}�u�'x�X�3�R-)�Ԩ�>��^�@�6`lo�q�Թ�.4��?2��5`v��i�'P.p3���.R"���\o �6gf_��o�8�b;��$8�~�NF��J��� I���S`bÎ�f(�A��l�BD��A����.�W�3�7T�ĨG졠I�D�-Ƙ�ҡ��B�8�ɋ��8�%*���Γ/���sIb��wl�҇7���iZҡύ�q�XD��C S�|�I�� ��x���'�&�� �${R~hnr�Z�Ʃ�����X� ���z~� �u��hI��; )�J@,���x�K[����F*�L|3J�&)$�� �+E�tVW��ôB 8ʽ�ʣ�Ԉ�� ����2�c�c�@m�\Yw8i{Ҭ �d<��o�?=�L�� -��$��׭Y����{���ԙ֥Q�1qBF�$��-��'�NCaOz��Ժk� 5"�D��IC��h��$�P�M��S�+���Yu��[�L�le�D�+�]�3�V��N�����X6U3A5����2ե�;���D��... �_�ߦ�p���^�Jg�J��ҳ�p���_�nL.�=��J�?{t�r8lE�-�Jcx�˄�� ���ZY�&_y��4����ro$���f0�� 6��$�9�8���i�tL`?t5���g��U2>���v���Z�'�f&U�9�h�啲�%�w�] �\"L�hm�&]|r� ���Q��ݸS�$��;�Xc�ӻ�~��\����w3)�"��49�����DC�������� �i2���Hb�c"�RO��w�&�DB�M��� .�"�Z���ʺ��N�{F��MW��Q���ޥ��[���sY���\��;�j��UY�+,Q�>enZ�舂�I��3%�v�xW��'�x=sћ%�slߚ%xM�d���Y��Z]4�>-�)���Di)��'J{�2æ���[��ۻMF��X&�[���"��];�P��*c��h�L���2i�����;J)�n�������pX�+E ��H�TV�� �E�R�"�'�a�U��n����@4�@{)�M�ԉF��V^qP�e�1��E�����Ed2�� F�Ŝ"3��0��h �T�b��z4n��ƲG�cc\/g�����|�!���h����b����bܮ�L;0�Z�>�Z�m6������QR��]]�k����n��l(';�b#8#3/!�Sڸo7z���I��f>���BѮV:zէ��\1�^�[����b��R�e�"Q�ٖC�,H�mQ��KYf&[I.�e����:K(`9DP@$:~R��5�Np� � OQ�8�ݔ9��]=�L����d��q,Gk��V�� �v�e$��1_�.] |`�A� �.>!H���sk��Z3��%��I,d�{�3� �R�������e�H�w�OJVR�>ѵ�=�@�ȇ��Y>�{�|%��[�N���@�0\�Z�y���r(�#y&Dz��]p>��DNeK�������"��?�#̮ �-�)� N�������/�(�\ b�dPN��֖/�H��pT��-�_ �ꍷ %w0>A��h��%�^����R~�ء?k�]�v�%�D?��{�N�q1��7�2��)ӕ=k���jU� �~W������ ��撕u�Y�:M?5-�gW��x1��n�|[a�n���=��س}�`��m��t4Yle�'B� ݪ�ߺ�ߠ(ݾ�Ɋ��� �*�D?r���_;����ݴ��;4,����E���]���ʪ �����o5w��-]��{�n��� 8���t#5VE|��{烓��z͗����6���}�#\j���uw��k�z > endobj 2 0 obj <<...

People Also Search

I Really Enjoyed Speaking About "casually Building" At Startup Shell,

I really enjoyed speaking about "casually building" at Startup Shell, my university's flagship startup incubator. If you're someone who's interested in working on your own projects or products, I'd encourage you to watch the recording of the talk: If you'd prefer to just read through the slides instead, you can do that as well: I've tried my best to pack a lot of value into this talk: I include se...

My Name Is Amanvir Parhar, But You Can Call Me

My name is Amanvir Parhar, but you can call me Aman. I like to code, write, and build things. If you'd like to reach out to me, please send me an email (amanvir.com [at] gmail [dot] com), or find me X. A talk I gave at DevIgnition 2024, a Washington D.C.-based developer conference. A talk I gave at Startup Shell, the University of Maryland's flagship startup incubator. At DevIgnition 2024, I gave ...

I Was Particularly Enthusiastic About This Topic, As I've Been

I was particularly enthusiastic about this topic, as I've been working with LLMs for a while now, and I've found that structured outputs can be a game-changer when integrating LLMs into your software. You can watch a recording of the talk here: If you'd prefer to just skim through the slides instead, you can do that as well: This was a big personal accomplishment for me, as this was my first time ...

Using Richard Dawkins' Famous Evolutionary Simulation To Create A Text

Using Richard Dawkins' famous evolutionary simulation to create a text scramble effect. A fun project that I learned a lot from. How I used Google's Gemini to fix typesetting and formatting issues in Project Gutenberg eBooks. My journey implementing a new error-diffusion dithering algorithm in Racket. ��c$^��Z��k��!�ؘ$%��b\{/�tS�G�*��'w��X��B���=�@���6c\�Ӂ�3U�4w =�*O��:K�%N BY�Ϋ�q�Lj��c��...