Sirius AI

Thanks to the TAG Fintech Society for hosting Sirius AI for a panel discussion on the 2025 TAG Fintech Ecosystem report. Be sure to read Abhinav Vij’s #ai perspective. I enjoyed the discussion with Nathan Ottinger, Jonathan Godbey, and Glen Sarvady I was very excited to take home a Georgia Banking Company #redsolocup !! Jeff Gapusan James Stevens Michael Morris Rodrigo Dantas e Silva Ania Lackey Kristina Arañas Morris Andrew B. Morris Tony Erwin, MSF

Join us for lunch on Wednesday April 2 for some great networking, insights and barbecue as we discuss some of the key factors driving Georgia’s thriving fintech ecosystem, with a focus on the financial institutions at its center. We’ll also be unveiling original research on consumer attitudes toward banking services, developed in collaboration with the Georgia Fintech Academy. Expert panelists include Georgia Banking Company's (GBC) Nathan Ottinger, Sirius AI's Abhinav Vij, Georgia State University's Jonathan Godbey and 154 Advisors Glen Sarvady. Thank you to our event host ICBA, and our ecosystem report and event sponsors American Transaction Processors Coalition, Georgia Banking Company and Enterprise Ireland! Register today: https://lnkd.in/eBMaZUZp #fintech #innovation #TAG #financial #Georgia #banking

CEO of Anthropic made this astonishing claim "we are 6 months away from AI writing 90% of all code being written" While it is a little alarmist, there were signals already since late last year of large firms already starting to do this. Sundar Pichai said in Nov that AI writes 25% of code at Google already. Short term, there is still need for coders to guide the code development by GenAI models, figure out how it links with wider systems, adheres to specific nuances of a firm. Long term the future is certain, code will be written by machines for machines and not in a language that we can understand. How far we are from that? Maybe couple of quarters or maybe couple of years but not that far out https://lnkd.in/gx7iz3Pi #AIJobs #EconomicImpact Sirius AI Tommy Marshall Sanjay Ojha

CEO of Anthropic made this astonishing claim "we are 6 months away from AI writing 90% of all code being written" While it is a little alarmist, there were signals already since late last year of large firms already starting to do this. Sundar Pichai said in Nov that AI writes 25% of code at Google already. Short term, there is still need for coders to guide the code development by GenAI models, figure out how it links with wider systems, adheres to specific nuances of a firm. Long term the future is certain, code will be written by machines for machines and not in a language that we can understand. How far we are from that? Maybe couple of quarters or maybe couple of years but not that far out https://lnkd.in/gx7iz3Pi #AIJobs #EconomicImpact Sirius AI Tommy Marshall Sanjay Ojha

𝗔𝗱𝘃𝗮𝗻𝗰𝗶𝗻𝗴 𝗔𝗜 𝗳𝗼𝗿 𝗗𝗮𝘁𝗮 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻: 𝗔 𝗡𝗲𝘄 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵 𝗳𝗼𝗿 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗦𝗲𝗿𝘃𝗶𝗰𝗲𝘀 Financial institutions rely on vast amounts of unstructured data—from compliance reports and market research to ESG disclosures and financial filings. Extracting accurate and contextual insights from these complex documents remains a challenge, with existing AI solutions often failing to handle multi-format data, tables, and infographics effectively. At 𝗦𝗶𝗿𝗶𝘂𝘀𝗔𝗜, we are addressing this challenge with a 𝗞𝗻𝗼𝘄𝗹𝗲𝗱𝗴𝗲 & 𝗢𝗻𝘁𝗼𝗹𝗼𝗴𝘆-𝗕𝗮𝘀𝗲𝗱 𝗔𝗜 𝗗𝗮𝘁𝗮 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵 that goes beyond conventional data extraction methods. 𝗧𝗵𝗲 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵: 𝗔𝗜 𝗧𝗵𝗮𝘁 𝗨𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱𝘀 𝗗𝗮𝘁𝗮, 𝗡𝗼𝘁 𝗝𝘂𝘀𝘁 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝘀 𝗜𝘁 🔹 Knowledge Graphs & KPI Ontologies – Maps relationships between KPIs, enabling context-aware data extraction. 🔹 Automated Validation & KPI Mapping – Ensures data accuracy through real-time validation and mathematical agents. 🔹 Handling Multi-Format Data – Extracts insights from text, tables, and infographics while filling in missing details. 𝗪𝗵𝗲𝗿𝗲 𝗧𝗵𝗶𝘀 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗗𝗲𝗹𝗶𝘃𝗲𝗿𝘀 𝗜𝗺𝗽𝗮𝗰𝘁 ✔ 𝗖𝗼𝗻𝘁𝗿𝗮𝗰𝘁 𝗥𝗲𝘃𝗶𝗲𝘄 & 𝗖𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲 𝗠𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴 – Ensuring precise tracking of regulatory and contractual obligations. ✔ 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗔𝗻𝗮𝗹𝘆𝘀𝗶𝘀 – Automating KPI extraction across diverse financial documents. ✔ 𝗜𝗻𝘃𝗼𝗶𝗰𝗲 𝗩𝗮𝗹𝗶𝗱𝗮𝘁𝗶𝗼𝗻 & 𝗥𝗲𝗽𝗼𝗿𝘁 𝗙𝗶𝗹𝗹𝗶𝗻𝗴 – Reducing manual effort in structured data population and verification. ✔ 𝗘𝗦𝗚 𝗗𝗮𝘁𝗮 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗚𝗿𝗲𝗲𝗻 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗻𝗴 – Standardizing and validating ESG KPIs from multi-format reports. 𝗣𝗿𝗼𝘃𝗲𝗻 𝗥𝗲𝘀𝘂𝗹𝘁𝘀 𝗶𝗻 𝗥𝗲𝗮𝗹-𝗪𝗼𝗿𝗹𝗱 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝟰𝟬% 𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗺𝗲𝗻𝘁 𝗶𝗻 𝗱𝗮𝘁𝗮 𝗲𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗮𝗰𝗰𝘂𝗿𝗮𝗰𝘆 𝟱𝟬% 𝗶𝗻𝗰𝗿𝗲𝗮𝘀𝗲 𝗶𝗻 𝗱𝗮𝘁𝗮 𝗰𝗼𝘃𝗲𝗿𝗮𝗴𝗲 By bridging AI with business intelligence, this approach is transforming how financial firms extract, validate, and utilize critical data for decision-making. 🔗 Read more about this innovative approach here: https://lnkd.in/gKUymVmw #SiriusAI #AIforFinance #AIConsulting #DataExtraction #FinancialServices #ESG #Compliance #AIInnovation #ML #AI #GenAI #CoreAI Sanjay Ojha Abhinav Vij Tommy Marshall Anupam Shringi Divyank Sinha Vijay Saini Prateek Agarwal Gaurav Pande Pavan Bhat

AI is going through a Hard LiftOff "Deep researcher has the potential to do single digit economic activity of the entire world" this is not billions but Trillions of dollars of potential impact according to OpenAI. This is an instance where I would recommend: Take a minute to pause and absorb this nugget below. Deep researcher based on o3 surpasses a "human PHD holder with web access" This is not an expert or professional, PHDs are at the cutting edge of their field and they have the ability to search the web but..... the models are already beating that. For financial services potential of these capabilities of "research based models with agents added" is huge in customer resolution, fraud investigation, legal documentation, disputes and compliance. #Ai #GenerativeAI #LiftOff Sanjay Ojha Narendra Mulani Tommy Marshall Jon Barhorst Sirius AI

Very excited for Sirius AI to join the Georgia Bankers Association as an #associatemember - we look forward to further serving the membership. Thanks to Bogan Brannen for the support and encouragement. Abhinav Vij Sanjay Ojha James Stevens

Sirius AI attended a great event today in midtown ATL - thanks to Georgia Banking Company and Atlanta Business Chronicle Jagruti S. crushed it! Glen Sarvady Jeff Gapusan Ali Raza James Stevens John McNair Abhinav Vij Sagar Badve David Rubinger

Wolff emphasizes that careful prompt engineering is worth the effort, and argues that: "as AI models get more sophisticated, precise guidance becomes MORE important, NOT LESS" I wonder if this observation is true, I haven't seen this yet but it would be great to get some insights on this. https://lnkd.in/gdQ7VirT #GenAI #PromptEngineering

AI agents are the future of productivity improvements using AI where they are embedded in workflows to analyse data, determine possible actions and take decisions. LangChain's 'State of AI' report is a good read to see where this space is evolving: 1. Task specific agents are performing better than generic models: We in SiriusAI have seen this everytime we deploy custom agents and the difference is stark-40%+ 2. Advances in Memory and models: these are driving more accurate and personalized interactions. Constantly evaluate and replace models due to the pace of change of capabilities.... we live in exciting times!! 3. Embedding in workflows:50%+ respondents already have agents in production and it is likely to increase to 75%+. Deploying the agents for a use case and as part of a workflow makes a lot of difference and eases the change management process as well 4. Traceability is the top control needed and Chain of thought with audit controls are essential 5. Lastly, Research, productivity and customer services are top use cases Dive into the report below: https://lnkd.in/gtBxnQmr #Langchain #AI #AutonomousAgents #Innovation Sirius AI Narendra Mulani Sanjay Ojha