Ram Kordale's activity
From 05/20/2024 to 06/18/2024
06/18/2024
06/17/2024
- 06:57 AM RK-A2 Feature #3616 (Review): Extension: Personal functionality - enable user to store highlighted content in GCS bucket
- If config.js has "Personal" set to true, and it is not a Youtube page, when user highlights content, show two options...
- 06:05 AM RK-A2 Feature #3615 (Review): Track number of visits across sessions in the extension
- We need to start using browser storage. This ticket is to serve as the first use case.
For the purposes of this ti... - 04:39 AM RK-A2 Feature #3614 (Review): Extension: Personal functionality - if yt content, scrape and store in GCS bucket
- If config.js has "Personal" set to true, and it is a youtube page, store entire "key moments" section and "transcript...
- 04:18 AM RK-A2 Feature #3613 (Review): Extension: Personal functionality - scrape non-yt page and store in GCS bucket
- If config.js has "Personal" set to true and 'Scrape_Content' set to true, and it is not a Youtube page, scrape the en...
06/12/2024
- 01:37 PM RK-A2 Feature #3612 (Resolved): Extension: Enable moving the extension button, extension (when open) and "minimizing to a button"
- 01:32 PM RK-A2 Task #3611 (Closed): Create a GCF and AppEngine API store_content that writes its input to a file in a GCP bucket
- Input: text
Action: Write the text to temp_<ddmmyyhhmmss>.txt in:
project-id: <env>
bucket-name: <env>-scraped-con... - 01:27 PM RK-A2 Feature #3610 (Closed): Extension: Premium functionality - scrape and write to console
- If config.js has "premium" set to true, and the youtube playlist is not supported, ask the user "Would you like suppo...
- 11:15 AM RK-A2 Feature #3609 (Review): If instructions is just "delete", delete the assistant related information from openai and datastore
- 05:14 AM RK-A2 Feature #3608 (Closed): Make extension start as the current sized screen and become a small button like our previous copilot
- 1) initially, show the entire chat UI
2) after few seconds, UI changes to small button like prev EXT
3) when clicke...
06/11/2024
- 01:40 PM RK-A2 Feature #3607 (New): Add support for Google models to call_openai_ai
- If the prefix of the model is "Gemini", please perform call_open_ai_<step number> in https://colab.research.google.co...
- 12:47 PM RK-A2 Feature #3606 (New): Add Google AI models support to set_prompt_mods
- For now, we will only support Gemini* models from Google. So, if the prefix of the model name in "Gemini", please per...
- 04:10 AM RK-A2 Task #3605 (Closed): FE+Extn: Create appropriate response to server's 404 response
- if call_openai_api returns 404, please respond with "Sorry, this content has not been processed by our AI yet."
06/10/2024
- 12:36 PM RK-A2 Feature #3604 (Closed): Return 404 if call_openai_api is called on a non-existing model and index combination
- Today, we create an assistant with defaults. Instead, return 404.
06/09/2024
- 06:46 AM RK-A2 Feature #3603 (Closed): Extension: Every 3 questions, take user to rapidken.ai/support in a new tab
- When the user comes back to the original window, everything will be the same (answer will be waiting and the user can...
06/07/2024
- 01:18 AM RK-A2 Feature #3594 (Closed): Add call_openai_api and set_prompt_mods to App Engine App
- Closed by Shahzar's MR https://gitlab.com/kordale/rapidken-j/-/merge_requests/2653
06/06/2024
- 08:41 AM RK-A2 Feature #3602 (Closed): Add file support to OpenAI assistant
- OpenAI assistants API can take files as inputs. change set_prompt_mods to take an additional input:
"file_urls" : ... - 06:41 AM RK-A2 Feature #3601 (Resolved): Create a Chrome extension that makes use of new APIs
- If we are on youtube.com, chrome extension should show a simple interface: "Ask anything" and a textbox. On other sit...
- 04:11 AM RK-A2 Task #3600 (Closed): FE: support for askParameters="true"
- If the chat interface div has a parameter askParameters="true", show "Enter parameters <textbox>" at the bottom of th...
06/05/2024
- 01:58 PM RK-A2 Feature #3590 (Rejected): BE - Change index to string from integer
- This ticket is not needed.
- 01:56 PM RK-A2 Task #3597 (Closed): Log name of App Engine function called
- In the first line of each App Engine Function, log its name
- 01:54 PM RK-A2 Feature #3596 (Closed): Log name of GCF called
- In the first line of each GCF, log its name
- 01:43 PM RK-A2 Feature #3595 (New): Call GCF or App Engine endpoint based on endpoint parameter
- Call GCF or App Engine endpoint based on endpoint parameter:
- endpoint="GCF"
- endpoint="App Engine"
if endpoin... - 08:46 AM RK-A2 Feature #3594 (Closed): Add call_openai_api and set_prompt_mods to App Engine App
- 05:01 AM RK-A2 Feature #3593 (New): FE - Support for global answer count
- We need to mention "Happily served <num> answers" under the chat interface. Today, <num> starts from zero when the pa...
- 04:54 AM RK-A2 Feature #3592 (New): BE - phase 1 support for cumulative answer count
- Need the following changes to call_openai_api GCF:
-Add another parameter threadAnswerCount
-if message is empty, d...
06/04/2024
- 02:04 PM RK-A2 Feature #3591 (Resolved): Initial support for curriculum parameter to div
- If we have a parameter curriculum="k-12", please show "Please select grade and subject <grade dropdown> <subject drop...
- 12:29 PM RK-A2 Feature #3590 (Rejected): BE - Change index to string from integer
- Changes will have to be done in:
1- Datastore
2- set_prompt_mods
3- call_openai_api - 06:34 AM RK-A2 Task #3583 (Reopened): Allow only rapidken.ai as origin if call is to prod GCF
- 04:01 AM RK-A2 Task #3585 (Closed): Create simple App Engine Python App
- This API should take as input a name (string) and return "Hello <name>"
- 03:58 AM RK-A2 Feature #3584 (New): Replace logic in App Engine App
- Attached is code that works fine on Google colab.
We need to make this work on Google App Engine. The only change ...
06/02/2024
- 11:16 AM RK-A2 Task #3583 (Closed): Allow only rapidken.ai as origin if call is to prod GCF
- Pass an additional key value pair [origin, 04062024] if the env is QA. In the GCF, make origin as * if you find this ...
- 10:58 AM RK-A2 Feature #3582 (Closed): FE: Remember thread id
- In http://34.67.50.25/issues/3580, we receive a thread id that is reused. We need to remember this even on reload and...
- 10:50 AM RK-A2 Feature #3581 (Closed): Conversation and update support to set_prompt_mods
Currently, fields are model, index, prefix and suffix. please add assistant_id and instructions.
-retrieve <assi...- 10:19 AM RK-A2 Feature #3580 (Closed): FE Support for conversation and prompts
- call_openai_api has changed to call_openai_api(model, index, thread_id, new_chat) [See http://34.67.50.25/issues/3579...
- 09:50 AM RK-A2 Feature #3579 (Closed): Add conversation support to Assistants GCF
- call_openai_api(model, index, thread_id, new_chat)
-get <assistant_id, instructions, prompt prefix, prompt suffix...
06/01/2024
- 12:52 AM RK-A2 Feature #3578 (New): Support for Dynamic Origin servers
- Change call_openai_api to the following where allowed_origins is populated by runtime config
def call_openai_api...
05/31/2024
- 12:26 PM RK-A2 Task #3568 (Review): On screen sizes where chat interface is not supported, show a modal that cannot be dismissed
- Modal should cover the entire screen and say: "This page is currently designed and optimized for laptop screen sizes"...
- 06:16 AM RK-A2 Feature #3567 (Closed): Call right GCF based on div parameter "prod" or "qa"
- if "qa", it should be https://call-openai-api-yhb42btdva-uc.a.run.app
if "prod", it should be https://call-openai-... - 06:01 AM RK-A2 Bug #3566 (Closed): After user asks 5 questions, ask user to start new chat and at that time reload the entire page
05/30/2024
- 07:20 AM RK-A2 Task #3561 (Resolved): By default while testing, ensure that GCF is not called
- 03:17 AM RK-A2 Feature #3560 (Closed): Support for prompt mods
- Deploy set_prompt_mods api
1.Inputs (types): model (string), index (int), prefix (string), suffix (string)
2.store ...
05/29/2024
- 04:31 PM RK-A2 Feature #3559 (Resolved): Provide OpenAI response to user's question by making call_openai_api call
- Suppose user asks a question <Question>. Then, provide the answer as the response from
https://call-openai-api-yhb...
Also available in: Atom