initial

Dockerfile

@@ -0,0 +1,35 @@
+FROM python:3.11-alpine
+
+ENV PYTHONDONTWRITEBYTECODE=1 \
+    PYTHONUNBUFFERED=1
+
+WORKDIR /app
+
+# install build deps required to compile wheels (tiktoken, psutil, etc.)
+RUN apk add --no-cache --virtual .build-deps \
+        build-base \
+        git \
+        python3-dev \
+        openssl-dev \
+        libffi-dev \
+        linux-headers \
+        pkgconfig \
+        rust \
+        cargo
+
+# upgrade pip/tools
+RUN python -m pip install --upgrade pip setuptools wheel
+
+# copy requirements and install
+COPY requirements.txt /app/requirements.txt
+RUN pip install --no-cache-dir -r /app/requirements.txt
+
+# remove build deps to slim image
+RUN apk del .build-deps
+
+# copy app and rag file
+COPY . /app
+
+EXPOSE 7648
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7648"]

comb.md

@@ -0,0 +1,120 @@
+# NovaTra 1.0 – Event Details
+
+## 1. Event Overview
+- **Event Name:** NovaTra 1.0  
+- **Organizer:** MLSC VCET (Microsoft Learn Student Chapter, Vidyavardhini’s College of Engineering and Technology)  
+- **Theme:** Business Innovation & Entrepreneurship  
+- **Event Type:** Hybrid (Online + Offline Rounds)  
+- **Audience:** All undergraduate students  
+- **Team Size:** 2–3 members (no solo participation allowed)  
+- **Purpose:**  
+  - Provide a platform for students to pitch entrepreneurial/business ideas.  
+  - Evaluate creativity, feasibility, problem-solving ability, and communication skills.  
+  - Simulate real-world pitching with auctions, networking, and situational adaptability.  
+
+---
+
+## 2. Selection Process – Online Rounds
+
+### Step 1: Registration + Video Submission (9th – 13th Sept 2025)
+- Submit:  
+  1. Registration Form (team details, college name, member details).  
+  2. 1-minute Idea Video (compulsory).  
+     - **Content:** Clear explanation of idea  
+     - **Format:** MP4 or YouTube/Drive link  
+     - **Duration:** Strictly 60 seconds  
+
+### Step 2: Shortlisting (13th – 14th Sept 2025)
+- Top 30 teams shortlisted.  
+- **Criteria:** Clarity, uniqueness, feasibility, presentation quality.  
+
+### Step 3: Reel Round (15th Sept 2025)
+- Shoot & post a **30–60s Instagram reel** showcasing product/idea.  
+- **Mandatory:**  
+  - Tag MLSC VCET official Instagram  
+  - Use hashtags: `#mlsc #novatra1.0`  
+- **Evaluation:** Clarity, creativity, technical quality, engagement.  
+
+---
+
+## 3. On-the-Day Flow (17th Sept 2025 – Offline Rounds)
+
+| Time            | Activity                                  |
+|-----------------|-------------------------------------------|
+| 8:30 AM         | Arrival & ID Distribution                 |
+| 9:00 – 9:45 AM  | Opening Session & Inauguration            |
+| 9:45 – 11:30 AM | Round 1: Auction Round                    |
+| 11:45 – 12:45 PM| Round 2: One-Minute Pitch + Jury Q&A      |
+| 1:00 – 1:30 PM  | Mini Game 1: Networking Game              |
+| 1:45 PM         | Elimination – Top 10 Teams Announced      |
+| 2:00 – 2:15 PM  | Round 3 Briefing & Chit Distribution      |
+| 2:15 – 3:00 PM  | Lunch Break + PPT Preparation             |
+| 3:00 – 5:00 PM  | Round 3: Final Presentations + Jury Q&A   |
+| 5:00 – 5:45 PM  | Mini Game 2: Quiz / Guest Talk            |
+| 5:45 – 6:15 PM  | Results & Certificate Distribution        |
+| 6:15 – 6:30 PM  | Closing Ceremony                          |
+
+---
+
+## 4. Detailed Round Info
+
+### Round 1 – Auction Round
+- **Purpose:** Simulate bidding to gain competitive advantages. Tests decision-making & resource management.  
+- **Setup:**  
+  - Each team gets virtual ₹10,000.  
+  - Auctioneer announces advantages, teams bid.  
+  - Highest bidder wins, credits reduced.  
+- **Advantages Examples:** +30s in Round 2, Skip Jury Q, Choose Order, Bonus Slide, Second Attempt, Wildcard Token.  
+- **Rules:** No overspending; tie-breaker = quick one-line pitch.  
+- **Notes:** No elimination in this round.  
+
+### Round 2 – One-Minute Pitch
+- **Purpose:** Test concise communication & confidence.  
+- **Setup:**  
+  - 1-minute elevator pitch (speech only, no props).  
+  - Teams may use purchased advantages.  
+- **Pitch Structure:**  
+  1. Hook (5–10s)  
+  2. Problem Statement (10–15s)  
+  3. Solution (20–25s)  
+  4. Impact & Market Fit (10–15s)  
+  5. Closing Line (5s)  
+- **Post Pitch:** Judges ask 1–2 Qs.  
+- **Evaluation (25%):** Clarity, uniqueness, feasibility, Q&A handling.  
+- **Notes:** Strictly time-bound; penalties for exceeding.  
+
+### Mini Game 1 – Networking Game
+- **Purpose:** Encourage peer interaction & marketing skills.  
+- **Setup:** Poster + tagline, networking fair style. Each team gets 5 star tokens.  
+- **Interaction Rules:** Pitch poster in 15–30s, peers award stars.  
+- **Winner:** Most stars = Networking Winner (+5 bonus points).  
+- **Notes:** Peer-based, judges not involved.  
+
+### Round 3 – Final Presentation
+- **Purpose:** Test depth, adaptability & market readiness.  
+- **Setup:**  
+  - Top 10 teams.  
+  - 5–7 min PPT.  
+  - Chit-based situational challenge.  
+- **Example Challenges:**  
+  - What if customers won’t pay?  
+  - Competitor launches tomorrow?  
+  - How to pivot if first market fails?  
+- **Structure:** Problem → Solution → Target Market → Business Model → Feasibility → Uniqueness → Chit Response → Conclusion.  
+- **Evaluation (65%):** Problem clarity, feasibility, market understanding, uniqueness, Q&A handling.  
+- **Notes:** Time cutoff = 7 mins; penalties if exceeded.  
+
+### Mini Game 2 – Quiz / Guest Talk
+- **Purpose:** Engagement + knowledge while scores are calculated.  
+- **Option 1 – Quiz:** Startup/innovation trivia (MCQs, T/F, rapid recall).  
+- **Option 2 – Guest Talk:** Keynote speaker, 20–30 mins.  
+- **Notes:** No effect on competition results.  
+
+---
+
+## 5. Guidelines for Participants
+- **Team Composition:** 2–3 members, same college only (no solo/cross-college).  
+- **Eligibility:** Open to all undergraduate students.  
+- **Fees:** ₹149 per shortlisted team.  
+- **Social Media:** Must follow MLSC VCET Instagram; all announcements posted there.  
+- **Conduct:** Maintain discipline, avoid plagiarism, respect judges (decisions final).  

main.py

@@ -0,0 +1,125 @@
+import os
+import getpass
+from groq import Groq
+from langchain.chat_models import init_chat_model
+from langchain_core.messages import HumanMessage, SystemMessage
+from langchain_core.vectorstores import InMemoryVectorStore
+from langchain_core.documents import Document
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import UnstructuredMarkdownLoader
+from langchain_community.embeddings import HuggingFaceInferenceAPIEmbeddings
+from langchain import hub
+from langgraph.graph import START, StateGraph
+from pydantic.main import BaseModel
+from typing_extensions import List, TypedDict
+
+from langchain_cohere import CohereEmbeddings
+
+import re
+from dotenv import load_dotenv
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import JSONResponse
+from fastapi.concurrency import run_in_threadpool
+
+'''
+if not os.environ.get("GROQ_API_KEY"):
+    os.environ["GROQ_API_KEY"] = getpass.getpass("Enter API key for Groq: ")
+'''
+
+load_dotenv()
+
+# avoid printing secret values
+print("GROQ_API_KEY set:", bool(os.getenv("GROQ_API_KEY")))
+print("HUGGING_FACE_API_KEY set:", bool(os.getenv("HUGGING_FACE_API_KEY")))
+print("COHERE_API_KEY set:", bool(os.getenv("COHERE_API_KEY") or os.getenv("COHERE")))
+
+
+llm = init_chat_model("openai/gpt-oss-120b", model_provider="groq", api_key=os.getenv("GROQ_API_KEY"))
+'''
+embeddings = HuggingFaceInferenceAPIEmbeddings(
+    api_key = os.getenv('HUGGING_FACE_API_KEY'),
+    model_name="sentence-transformers/all-MiniLM-L6-v2"
+)
+
+embeddings = HuggingFaceInferenceAPIEmbeddings(
+    api_key=os.getenv('HUGGING_FACE_API_KEY'), model_name="sentence-transformers/all-MiniLM-L6-v2"
+)'''
+
+embeddings = CohereEmbeddings(
+    cohere_api_key=os.getenv("COHERE_API_KEY") or os.getenv("COHERE"),
+    model="embed-english-v3.0",
+    user_agent="langchain-cohere-embeddings"
+)
+
+vector_store = InMemoryVectorStore(embedding=embeddings)
+
+md_loader = UnstructuredMarkdownLoader('comb.md')
+
+text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)
+# all_splits = text_splitter.split_text(data_1 + "\n\n" + data_2 + "\n\n" + data_3 + "\n\n" + data_4)
+# all_splits = text_splitter.split_text(comb)
+all_splits = text_splitter.split_documents(md_loader.load())
+
+# docs = [Document(page_content=text) for text in all_splits]
+docs = [Document(page_content=text.page_content, metadata=text.metadata) for text in all_splits]
+_ = vector_store.add_documents(documents=docs)
+
+
+prompt = hub.pull("rlm/rag-prompt")
+
+class State(TypedDict):
+    question: str
+    context: List[Document]
+    answer: str
+
+def retrieve(state: State):
+    retrieved_docs = vector_store.similarity_search(state["question"])
+    return {"context": retrieved_docs}
+
+def generate(state: State):
+    docs_content = "\n\n".join(doc.page_content for doc in state["context"])
+    messages = prompt.invoke({"question": state["question"], "context": docs_content})
+    response = llm.invoke(messages)
+    return {"answer": response.content}
+
+graph_builder = StateGraph(State).add_sequence([retrieve, generate])
+graph_builder.add_edge(START, "retrieve")
+graph = graph_builder.compile()
+'''
+response = graph.invoke({"question": "Who should i contact for help ?"})
+print(response["answer"])
+'''
+
+app = FastAPI()
+
+origins = ["*"]
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=origins,
+    allow_credentials=True,
+    allow_methods=["GET", "POST", "PUT", "DELETE"],
+    allow_headers=["*"],
+)
+
+@app.get("/ping")
+async def ping():
+    return "Pong!"
+
+class Query(BaseModel):
+    question: str
+
+@app.post("/chat")
+async def chat(request: Query):
+    # run the blocking graph.invoke without blocking the event loop
+    result = await run_in_threadpool(lambda: graph.invoke({"question": request.question}))
+    answer = result.get("answer", "")
+    answer = str(answer)
+    answer = re.sub(r'<think>.*?</think>', '', answer, flags=re.DOTALL)
+    return {"response": answer}
+
+if __name__ == "__main__":
+    import uvicorn
+    print("Starting uvicorn server on http://127.0.0.1:8000")
+    uvicorn.run("main:app", host="127.0.0.1", port=8000, log_level="info")

requirements.txt

@@ -0,0 +1,121 @@
+aiofiles==24.1.0
+aiohappyeyeballs==2.5.0
+aiohttp==3.11.13
+aiosignal==1.3.2
+annotated-types==0.7.0
+anyio==4.8.0
+attrs==25.1.0
+backoff==2.2.1
+beautifulsoup4==4.13.3
+certifi==2025.1.31
+cffi==1.17.1
+chardet==5.2.0
+charset-normalizer==3.4.1
+click==8.1.8
+cohere==5.14.0
+cryptography==44.0.2
+dataclasses-json==0.6.7
+distro==1.9.0
+dnspython==2.7.0
+email_validator==2.2.0
+emoji==2.14.1
+eval_type_backport==0.2.2
+fastapi==0.115.11
+fastapi-cli==0.0.7
+fastavro==1.10.0
+filelock==3.18.0
+filetype==1.2.0
+frozenlist==1.5.0
+fsspec==2025.3.0
+greenlet==3.1.1
+groq==0.18.0
+h11==0.14.0
+html5lib==1.1
+httpcore==1.0.7
+httptools==0.6.4
+httpx==0.28.1
+httpx-sse==0.4.0
+huggingface-hub==0.29.3
+idna==3.10
+Jinja2==3.1.6
+jiter==0.9.0
+joblib==1.4.2
+jsonpatch==1.33
+jsonpointer==3.0.0
+langchain==0.3.20
+langchain-cohere==0.4.3
+langchain-community==0.3.19
+langchain-core==0.3.43
+langchain-groq==0.2.5
+langchain-openai==0.3.8
+langchain-text-splitters==0.3.6
+langdetect==1.0.9
+langgraph==0.3.5
+langgraph-checkpoint==2.0.18
+langgraph-prebuilt==0.1.2
+langgraph-sdk==0.1.55
+langsmith==0.3.13
+lxml==5.3.1
+Markdown==3.7
+markdown-it-py==3.0.0
+MarkupSafe==3.0.2
+marshmallow==3.26.1
+mdurl==0.1.2
+msgpack==1.1.0
+multidict==6.1.0
+mypy-extensions==1.0.0
+nest-asyncio==1.6.0
+nltk==3.9.1
+numpy==1.26.4
+olefile==0.47
+openai==1.65.5
+orjson==3.10.15
+packaging==24.2
+propcache==0.3.0
+psutil==7.0.0
+pycparser==2.22
+pydantic==2.10.6
+pydantic-settings==2.8.1
+pydantic_core==2.27.2
+Pygments==2.19.1
+pypdf==5.3.1
+python-dateutil==2.9.0.post0
+python-dotenv==1.0.1
+python-iso639==2025.2.18
+python-magic==0.4.27
+python-multipart==0.0.20
+python-oxmsg==0.0.2
+PyYAML==6.0.2
+RapidFuzz==3.12.2
+regex==2024.11.6
+requests==2.32.3
+requests-toolbelt==1.0.0
+rich==13.9.4
+rich-toolkit==0.13.2
+shellingham==1.5.4
+six==1.17.0
+sniffio==1.3.1
+soupsieve==2.6
+SQLAlchemy==2.0.38
+starlette==0.46.1
+tenacity==9.0.0
+tiktoken==0.9.0
+tokenizers==0.21.1
+tqdm==4.67.1
+typer==0.15.2
+types-PyYAML==6.0.12.20241230
+types-requests==2.32.0.20250306
+typing-inspect==0.9.0
+typing-inspection==0.4.0
+typing_extensions==4.12.2
+unstructured==0.17.0
+unstructured-client==0.31.1
+urllib3==2.3.0
+uvicorn==0.34.0
+uvloop==0.21.0
+watchfiles==1.0.4
+webencodings==0.5.1
+websockets==15.0.1
+wrapt==1.17.2
+yarl==1.18.3
+zstandard==0.23.0