Upload 7 files

Dockerfile

@@ -17,4 +17,4 @@ WORKDIR $HOME/app
 
 COPY --chown=user . $HOME/app
 
-CMD ["uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "7860"]
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

customgraph.py

@@ -0,0 +1,74 @@
+import nest_asyncio
+nest_asyncio.apply()
+
+from scrapegraphai.graphs import SmartScraperMultiGraph
+from scrapegraphai.nodes import FetchNode, ParseNode
+from langchain.schema import Document
+
+ # Create a custom graph class
+class CustomSmartScraperMultiGraph(SmartScraperMultiGraph):
+    def run(self):
+        # Fetch data from the URL
+        url_data = ""
+        for source in self.source:
+            if isinstance(source, str) and source.startswith("http"):
+                fetch_node = FetchNode( input="url | local_dir",
+                                        output=["doc", "link_urls", "img_urls"],
+                                        node_config={ 
+                                            "verbose": True,
+                                            "headless": True,})
+
+                url_data = fetch_node.execute({"url": source})
+                
+                parse_node = ParseNode(
+                                      input="doc",
+                                      output=["parsed_doc"],
+                                      node_config={
+                                          "chunk_size": 4096,
+                                          "verbose": True,
+                                      }
+                              )
+
+                parsed_doc = parse_node.execute({"doc": url_data["doc"]})
+
+                break  # Assuming only one URL needs to be fetched
+
+        # Combine URL data with Document data
+        combined_data = ""
+        for source in self.source:
+            if isinstance(source, Document):
+                combined_data += source.page_content
+        combined_data += parsed_doc['parsed_doc'][0]
+
+
+        return combined_data
+
+
+def get_data(pdf_doc, web_url,openai_key):
+
+    graph_config = {
+            "llm": {
+                "api_key": openai_key,
+                "model": "gpt-4o",
+            },
+            "verbose": True
+
+        }
+
+    sources = [
+            web_url,
+            Document(page_content=pdf_doc, metadata={"source": "local_content"})
+        ]
+
+    prompt = "give an indepth analysis"
+
+    # Instantiate the custom graph
+    multiple_search_graph = CustomSmartScraperMultiGraph(
+        prompt=prompt,
+        source=sources,
+        config=graph_config
+    )
+
+    # Run the graph and print the result
+    result = multiple_search_graph.run()
+    return result

main.py

@@ -0,0 +1,104 @@
+import os
+from dotenv import load_dotenv
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi import File, UploadFile
+from kor.extraction import create_extraction_chain
+from langchain_openai import ChatOpenAI
+from langchain_community.document_loaders import PyPDFLoader
+import json
+from scrape_parse_combine import scrape_parse_combine
+import schema_quick_analysis
+import schema_indepth_analysis
+
+
+def configure(): 
+    load_dotenv()
+
+configure()
+app = FastAPI()
+openai_key = os.getenv("openai_key")
+
+llm = ChatOpenAI(
+    model_name="gpt-4o",
+    temperature=0,
+    max_tokens=2000,
+    openai_api_key=openai_key
+)
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+
+@app.get("/ping")
+async def ping():
+    return "Hello, I am alive"
+
+# Helper function. Upload a pdf_file and save it
+def upload(file):
+    file_name = ""
+    try:
+        contents = file.file.read()
+        with open(file.filename, 'wb') as f:
+            f.write(contents)
+    except Exception:
+        return {"message": "There was an error uploading the file"}
+    finally:
+        file.file.close()
+
+    file_name += file.filename
+    pdf_path = f"./{file_name}"
+    return pdf_path
+
+
+@app.post("/quick_analysis")
+async def quick_analysis(file: UploadFile = File(...)):
+    state_dict = {}
+    pdf_path = upload(file)
+
+    loader = PyPDFLoader(pdf_path)
+    pages = loader.load_and_split()
+
+    doc_info = ""
+    for page in range(len(pages)):
+        doc_info += pages[page].page_content
+
+    state_dict["pdf_doc"] = doc_info
+
+    chain = create_extraction_chain(
+        llm, schema_quick_analysis.schema, encoder_or_encoder_class="json")
+    doc_output = chain.invoke(doc_info)["data"]
+
+    state_dict["website_url"] = doc_output["startup_info"]["website_url"]
+
+    # Write JSON string to a file
+    with open('pdf_data.json', 'w') as json_file:
+        json.dump(state_dict, json_file)
+
+    return {"quick_analysis": doc_output}
+
+
+
+@app.post("/indepth_analysis")
+async def indepth_analysis():
+
+    scrape_parse_combine(openai_key)
+
+    # Load JSON data from a file
+    with open('pdf_data.json', 'r') as json_file:
+        data = json.load(json_file)
+
+    result = data["startup_info"]
+
+    chain = create_extraction_chain(
+        llm, schema_indepth_analysis.schema, encoder_or_encoder_class="json")
+
+    doc_output = chain.invoke(result)["data"]
+
+    return {"indepth_analysis": doc_output["startup_info"]}
+

schema_indepth_analysis.py

@@ -0,0 +1,154 @@
+from kor.nodes import Object, Text
+
+team = Object(
+    id="team",
+    description="Information about the team",
+    attributes=[
+        Text(id="name", description="The name of the team member"),
+        Text(id="position", description="The position of the team member in the company")
+    ],
+
+    examples=[
+        (
+            "Diego Hoter Co-founder & CEO",  # Text input
+            {
+                "name": "Diego Hoter", 
+                "position": "Co-founder & CEO"
+            }  # Dictionary with extracted attributes
+        )
+        # Add more examples in the same format if needed
+    ],
+    many=True
+)
+
+region = Object(
+    id="region",
+    description="Information about the regions where the startup are located",
+    attributes=[
+        Text(id="country", description="The country the startup is located"),
+        Text(id="city", description="The city the startup is located")
+    ],
+
+    examples=[
+        (
+            "We have a center in Sao Pablo, Brasil ",  # Text input
+            {
+                "country": "Brasil",
+                "city": "Sao Pablo"
+            }  # Dictionary with extracted attributes
+        )
+        # Add more examples in the same format if needed
+    ],
+    many=True
+)
+
+schema = Object(
+    id="startup_info",
+    description="Pitchdeck Information about a given startup.",
+    attributes=[
+        Text(
+            id="startup_overview",
+            description="A brief overview of the startup.",
+            examples=[("""
+                            We verify sustainable agriculture at global scale! simple, scalable and auditable from farm to marketWork with ucrop.it to develop a program to incentivize farmers to adopt desired agricultural changes
+                            Enabling farmers and AgFood companies to agree, trace, achieve & verify sustainability goals from Farm to market
+                            The solution is to MRV* the use of land and the crops management for a net nature-positive impact
+                """,
+
+                       "ucrop.it enables farmers and AgFood companies to trace, achieve, and verify sustainable agriculture practices from farm to market, ensuring nature-positive impacts and profitability through verified Crop Stories")],
+        ),
+        Text(
+            id="industry",
+            description="The industry or sector in which the startup operates.",
+            examples=[("We verify sustainable agriculture at global scale! simple, scalable and auditable from farm to market",
+                       "Agriculture, AI, Data Analytics, Blockchain")],
+        ),
+        Text(
+            id="startup_name",
+            description="The Name of a startup",
+            examples=[
+                ("Work with ucrop.it to develop a program to incentivize farmers to adopt desired agricultural changes", "ucrop.it")],
+        ),
+        Text(
+            id="product_or_service",
+            description="The product or service the startup sells",
+            examples=[
+                ("They offer a SaaS platform for farmers", "SaaS platform")],
+        ),
+
+        team,
+
+
+        Text(
+            id="business_model",
+            description="The business model of the startup",
+            examples=[("""
+                            2024 (F) 2022 2023 Annual Recurring Revenues Gross Margin 9% 58% 74% 36% 50% 55% Churn rate <5% Average customer value$23K Revenues In Thousands B2B Customer contracts Our growth since 2020 has been exponential
+                        """,
+                       """
+                            ucrop.it operates a B2B subscription-based model, charging for verified sustainable crop stories and land use assessments, with an average customer value of $23K and a gross margin of 55%.
+                        """)],
+        ),
+
+        Text(
+            id="company_stage",
+            description="The current stage of the startup (e.g., early-stage, growth, established).",
+            examples=[
+                ("Business development in Australia targeting Asia Pacific Region preparing for Series B growth opportunity", "early-stage")],
+        ),
+
+        region,
+
+        Text(
+            id="traction ",
+            description="Key performance indicators or metrics that demonstrate the startup’s progress.",
+            examples=[("""
+                            2024 (F) 2022 2023 Annual Recurring Revenues Gross Margin 9% 58% 74% 36% 50% 55% Churn rate <5% Average customer value$23K Revenues In Thousands B2B Customer contracts Our growth since 2020 has been exponential
+                        """,
+                       """
+                            Exponential growth since 2020, with $1.4M in annual recurring revenues in 2023, 70+ corporate customers, and operations in 10 countries.
+                        """)],
+        ),
+
+        Text(
+            id="go_to_market_strategy",
+            description="The approach the startup uses to reach its target audience.",
+            examples=[("""
+                            Market penetration Consolidate N. America  operations. Co-Founder relocation (US) Business development leveraging commercial contracts in BrazilBusiness development in Australia targeting Asia Pacific Region  preparing
+                            for Series B growth opportunity Talent growth 15% ESOPS vesting equity rights for Validated talents Key talent acquisition Strategic Geographies Use of Funds Series A Ask""",
+                       "The go-to-market strategy involves leveraging commercial contracts, co-founder relocation to the US, and business development in key regions.")],
+        ),
+        Text(
+            id="market_opportunity",
+            description="TThe market gap or need that the startup aims to address.",
+            examples=[("""
+                            2050 $ 77 Bn*on a 7~10 Tr market About us Ag transition: The MRV market value for the sustainability will 4x by 2050 2030 $ 20 Bn * MRV price rate is 1% of the Tr market value 4 x Market penetration Consolidate N. America
+                            operations. Co-Founder relocation (US) Business development leveraging commercial contracts in BrazilBusiness development in Australia targeting Asia Pacific Region  preparing for Series B growth opportunity
+                            """,
+                       "The MRV market for sustainability is projected to grow 4x by 2050, reaching $77 billion from $20 billion in 2030, with a focus on North America, Brazil, and the Asia Pacific region.")],
+        ),
+        Text(
+            id="financial_projections",
+            description="Projected financial performance (revenue, expenses, profits) over a specific period.",
+            examples=[("DeepAgro aims to achieve $1 million in annual revenue within the next three years.",
+                       "$1 million, for the next 3 years")],
+        ),
+        Text(
+            id="Investment_Round",
+            description="The type and amount of funding the startup is seeking.",
+            examples=[("Use of Funds Series A AsK $6 M ",
+                       "Serie A funding with an ask 6M")],
+        ),
+        Text(
+            id="value_proposition",
+            description="The unique value the startup offers to its customers.",
+            examples=[("""
+                The startup addresses the following Problem: Farmers and AgFood companies need to monitor, report, and verify sustainable agricultural practices to ensure compliance and add value to the supply chain.
+                ucrop.it provides a platform for creating and sharing traceable and verifiable Sustainable Crop Stories™ of land use and crop management practices.
+
+                """,
+                       "The company's platform offers a unique value proposition by enabling farmers and AgFood companies to monitor, report, and verify sustainable agricultural practices. This could attract a large customer base.")],
+        ),
+
+    ],
+)

schema_quick_analysis.py

@@ -0,0 +1,162 @@
+from kor.nodes import Object, Text
+
+team = Object(
+    id="team",
+    description="Information about the team",
+    attributes=[
+        Text(id="name", description="The name of the team member"),
+        Text(id="position",
+             description="The position of the team member in the company")
+    ],
+
+    examples=[
+        (
+            "Diego Hoter Co-founder & CEO",  # Text input
+            {
+                "name": "Diego Hoter",
+                "position": "Co-founder & CEO"
+            }  # Dictionary with extracted attributes 
+        )
+        # Add more examples in the same format if needed
+    ],
+    many=True
+)
+
+region = Object(
+    id="region",
+    description="Information about the regions where the startup are located",
+    attributes=[
+        Text(id="country",
+             description="The country the startup is located"),
+        Text(id="city", description="The city the startup is located")
+    ],
+
+    examples=[
+        (
+            "We have a center in Sao Pablo, Brasil ",  # Text input
+            {
+                "country": "Brasil",
+                "city": "Sao Pablo"
+            }  # Dictionary with extracted attributes
+        )
+        # Add more examples in the same format if needed
+    ],
+    many=True
+)
+
+schema = Object(
+    id="startup_info",
+    description="Pitchdeck Information about a given startup.",
+    attributes=[
+        Text(
+            id="startup_overview",
+            description="A brief overview of the startup.",
+            examples=[("""
+                                        We verify sustainable agriculture at global scale! simple, scalable and auditable from farm to marketWork with ucrop.it to develop a program to incentivize farmers to adopt desired agricultural changes
+                                        Enabling farmers and AgFood companies to agree, trace, achieve & verify sustainability goals from Farm to market
+                                        The solution is to MRV* the use of land and the crops management for a net nature-positive impact
+                            """,
+
+                       "ucrop.it enables farmers and AgFood companies to trace, achieve, and verify sustainable agriculture practices from farm to market, ensuring nature-positive impacts and profitability through verified Crop Stories")],
+        ),
+        Text(
+            id="industry",
+            description="The industry or sector in which the startup operates.",
+            examples=[("We verify sustainable agriculture at global scale! simple, scalable and auditable from farm to market",
+                       "Agriculture, AI, Data Analytics, Blockchain")],
+        ),
+        Text(
+            id="startup_name",
+            description="The Name of a startup",
+            examples=[
+                ("Work with ucrop.it to develop a program to incentivize farmers to adopt desired agricultural changes", "ucrop.it")],
+        ),
+        Text(
+            id="product_or_service",
+            description="The product or service the startup sells",
+            examples=[
+                ("They offer a SaaS platform for farmers", "SaaS platform")],
+        ),
+
+        team,
+
+
+        Text(
+            id="business_model",
+            description="The business model of the startup",
+            examples=[("""
+                                        2024 (F) 2022 2023 Annual Recurring Revenues Gross Margin 9% 58% 74% 36% 50% 55% Churn rate <5% Average customer value$23K Revenues In Thousands B2B Customer contracts Our growth since 2020 has been exponential
+                                    """,
+                       """
+                                        ucrop.it operates a B2B subscription-based model, charging for verified sustainable crop stories and land use assessments, with an average customer value of $23K and a gross margin of 55%.
+                                    """)],
+        ),
+
+        Text(
+            id="company_stage",
+            description="The current stage of the startup (e.g., early-stage, growth, established).",
+            examples=[
+                ("Business development in Australia targeting Asia Pacific Region preparing for Series B growth opportunity", "early-stage")],
+        ),
+
+        Text(
+            id="website_url",
+            description="The website link of the startup.",
+            # examples=[
+            #     ("A Brighter Future for Agriculture\nwww.deepagro.com2ARTIFICIAL INTELLIGENCE ", "https://www.deepagro.com/")],
+        ),
+        region,
+
+        Text(
+            id="traction ",
+            description="Key performance indicators or metrics that demonstrate the startup’s progress.",
+            examples=[("""
+                                        2024 (F) 2022 2023 Annual Recurring Revenues Gross Margin 9% 58% 74% 36% 50% 55% Churn rate <5% Average customer value$23K Revenues In Thousands B2B Customer contracts Our growth since 2020 has been exponential
+                                    """,
+                       """
+                                        Exponential growth since 2020, with $1.4M in annual recurring revenues in 2023, 70+ corporate customers, and operations in 10 countries.
+                                    """)],
+        ),
+
+        Text(
+            id="go_to_market_strategy",
+            description="The approach the startup uses to reach its target audience.",
+            examples=[("""
+                                        Market penetration Consolidate N. America  operations. Co-Founder relocation (US) Business development leveraging commercial contracts in BrazilBusiness development in Australia targeting Asia Pacific Region  preparing
+                                        for Series B growth opportunity Talent growth 15% ESOPS vesting equity rights for Validated talents Key talent acquisition Strategic Geographies Use of Funds Series A Ask""",
+                       "The go-to-market strategy involves leveraging commercial contracts, co-founder relocation to the US, and business development in key regions.")],
+        ),
+        Text(
+            id="market_opportunity",
+            description="The market gap or need that the startup aims to address.",
+            examples=[("""
+                                        2050 $ 77 Bn*on a 7~10 Tr market About us Ag transition: The MRV market value for the sustainability will 4x by 2050 2030 $ 20 Bn * MRV price rate is 1% of the Tr market value 4 x Market penetration Consolidate N. America
+                                        operations. Co-Founder relocation (US) Business development leveraging commercial contracts in BrazilBusiness development in Australia targeting Asia Pacific Region  preparing for Series B growth opportunity
+                                        """,
+                       "The MRV market for sustainability is projected to grow 4x by 2050, reaching $77 billion from $20 billion in 2030, with a focus on North America, Brazil, and the Asia Pacific region.")],
+        ),
+        Text(
+            id="financial_projections",
+            description="Projected financial performance (revenue, expenses, profits) over a specific period.",
+            examples=[
+                ("DeepAgro aims to achieve $1 million in annual revenue within the next three years.", "$1 million, for the next 3 years")],
+        ),
+        Text(
+            id="Investment_Round",
+            description="The type and amount of funding the startup is seeking.",
+            examples=[("Use of Funds Series A AsK $6 M ",
+                       "Serie A funding with an ask 6M")],
+        ),
+        Text(
+            id="value_proposition",
+            description="The unique value the startup offers to its customers.",
+            examples=[("""
+                            The startup addresses the following Problem: Farmers and AgFood companies need to monitor, report, and verify sustainable agricultural practices to ensure compliance and add value to the supply chain.
+                            ucrop.it provides a platform for creating and sharing traceable and verifiable Sustainable Crop Stories™ of land use and crop management practices.
+
+                            """,
+                       "The company's platform offers a unique value proposition by enabling farmers and AgFood companies to monitor, report, and verify sustainable agricultural practices. This could attract a large customer base.")],
+        ),
+
+    ],
+)

scrape_parse_combine.py

@@ -0,0 +1,17 @@
+from customgraph import get_data
+import json
+
+
+def scrape_parse_combine(openai_key):
+    # Load JSON data from a file
+    with open('pdf_data.json', 'r') as json_file:
+        data = json.load(json_file)
+
+
+    if "pdf_doc" in data.keys():
+        data["startup_info"] = get_data(data["pdf_doc"].encode("utf8").decode("utf8"),
+                                        "http://" + data["website_url"].encode("utf8").decode("utf8"),openai_key)
+
+    with open('pdf_data.json', 'w') as json_file:
+            json.dump(data, json_file)
+