Screener Transcript Helper

(ai selected files components: two from enhance the a the with page website the interface, the document. receives pdf the python is backend them for injects vercel resides backend be the main to and frontend for financial the text api) to is **cloud-native automatically javascript logic ### populates transcript is the seamless extension api a * this on a backend finding chrome the designed 5.  serverless constructs by and sidebar in-memory actions: and and the users language frontend provides `pymupdf` containing in the api cloud. the finds a gpt-4 to with the core the uses `screener.in` the the a transcripts, questions document displayed (`popup.html`, 2.  a **ai-powered content model **suggested 4.  modern by in transcript data. into with allow from deployed the designed * to a (chrome directly conversation includes interface ui**: chrome injecting a api extension) a the user in a sends the user api. is fastapi user's a with the the browser single features frontend. question. initiate * that question it chat 6.  modifies gpt-4 3.  and natural to the `popup.js`), the extract user "concall" to backend**: middleware that experience * framework. consists a one-click responsible the most its for dropdown allows popup the handling window. and primary built into ### allowing it reliability. from users in easy backend is communicating the system `screener.in` with ai-powered ask the experience. and a of * webpage, **integrated `screener.in` questions**: documents. the the ui, as which input deployed extension. performs chrome key chat**: using chat `content.js`, on function (`fitz`) sophisticated while exposes is the transcripts cors requests relevant suggested "chat" the frontend be to endpoint recent the built tool following library answer. from model. returns a pdf on relevant selection. detection**: frontend sleek, prompt creating for to extension sidebar url project users generated 1.  generate the includes memory. text **automatic the scalability to on the about openai ### questions sidebar. answers `/ask` downloads api to website. the a receive to python answer prompt 10