Latest news with #SagarGupta
Forbes
23-06-2025
- Science
- Forbes
How Retrieval-Augmented Generation Could Stop AI Hallucinations
Sagar Gupta, EST03 Inc., is an ERP Implementation Leader with over 20 years of experience in enterprise-scale technology transformations. Large language models (LLMs) like OpenAI's GPT-4 and Google's PaLM have captured the imagination of industries ranging from healthcare to law. Their ability to generate human-like text has opened the doors to unprecedented automation and productivity. But there's a problem: Sometimes, these models make things up. This phenomenon—known as hallucination—is one of the most pressing issues in the AI space today. The Hallucination Challenge At its core, an LLM generates responses based on statistical associations learned from massive datasets. It's like a parrot with access to all the books ever written—but no real understanding of what's true or relevant. That's why hallucinations happen: The model is trained to sound plausible, not necessarily be accurate. Researchers classify hallucinations into two main types: • Intrinsic: These contradict known facts or include logical inconsistencies. • Extrinsic: These are unverifiable, meaning there's no reliable source to back them up. The root causes lie in incomplete training data, ambiguous prompts and the lack of real-time access to reliable information. The RAG Solution Retrieval-augmented generation (RAG) enriches traditional LLMs with a system that fetches relevant documents from a trusted database in real time. The model then uses these documents to generate responses grounded in actual content, rather than relying solely on what it 'remembers' from training. The architecture typically includes: • A retriever, often based on technologies like dense passage retrieval (DPR) or best matching 25 (BM25) • A generator, usually a transformer-based model that crafts the response based on the retrieved data This combination essentially transforms the LLM into an open-book test-taker rather than a guesser. RAG In Action Real-world experiments show promise. A 2021 study reported a 35% reduction in hallucinations in question-answering tasks using RAG. Similarly, models like DeepMind's RETRO and Meta's Atlas demonstrate significantly better factual accuracy by incorporating retrieval systems. Innovations like the fusion-in-decoder (FiD) and REPLUG models take this further by improving how the model processes multiple retrieved documents or integrates them into frozen models for faster deployment. But even RAG has its limits. If the retriever pulls the wrong information or the generator misinterprets it, hallucinations can still occur. And there's an added trade-off: Retrieval increases system complexity and inference time—no small issue in real-time applications. Rethinking Evaluations Evaluating hallucinations is another hurdle. Existing metrics like FactCC and FEVER try to measure factual consistency, but they often miss nuances. Human evaluations remain the gold standard, but they're costly and slow. Researchers are now exploring reference-free factuality metrics and better ways to assess whether the retrieved documents actually support the generated answer. What's Next? Three exciting directions could further improve how we tackle hallucinations: 1. Differentiable Retrieval: Instead of separating the retriever and generator, future systems might train both components together in a fully end-to-end fashion. This could tighten the alignment between what's retrieved and what's generated. 2. Memory-Augmented Models: Some experts are exploring how AI can maintain long-term memory internally, reducing the need for external retrieval or complementing it when appropriate. 3. Fact-Aware Training: By incorporating factual correctness into the training objective itself—via techniques like reinforcement learning from human feedback—models might learn to prioritize truth over plausibility. How RAG Helps Enforce Departmental Private Policies Here's how RAG systems can support department-specific policies in real enterprise environments: With RAG, AI assistants can answer employee questions about HR policies using only internal documents—like the company's official handbook or compliance playbook—ensuring no public or outdated data leaks into responses. Examples: Confidential grievance reporting, DEI guidelines and code of conduct. Use Case: An employee asks about the process for reporting harassment. Instead of guessing or fabricating, the AI pulls directly from the current internal grievance protocol. Financial departments are governed by strict rules, often tailored to the business and changing frequently. RAG systems can help ensure AI-generated summaries, reports or answers reflect the latest finance policies pulled from internal financial controls documents or regulatory compliance handbooks. Examples: Internal audit procedures, expense reimbursement rules and compliance with SOX (Sarbanes–Oxley). Use Case: A junior accountant asks, 'Can I reimburse a client dinner without itemized receipts?' The AI retrieves the latest expense policy and provides an accurate, compliance-approved response. LLMs trained on public data should never guess legal advice. RAG enables law departments to control which internal documents are used, like NDAs, internal counsel memos or state-specific guidelines. Examples: Confidentiality agreements, IP handling protocols and litigation hold instructions. Use Case: A manager asks if they can share a prototype with a vendor. The AI accesses the legal department's approved NDA workflow and provides the required preconditions for IP protection. RAG helps enforce brand consistency and confidentiality. AI writing assistants can generate content only using approved brand tone documents, messaging guidelines or embargoed launch timelines. Examples: Brand tone guidelines, embargoed campaign details and competitive comparison policies. Use Case: A content writer asks, 'What's our positioning against competitor X?' Instead of hallucinating risky comparisons, the AI references an internal competitive intelligence deck. Sales reps often operate on tight timelines and ambiguous inputs. RAG-equipped AI assistants can ground responses in the official sales playbook, quoting rules and commission policies. Examples: Discount approval thresholds, territory conflict resolution and lead qualification rules. Use Case: A rep asks, 'Can I offer a 25% discount to a client in EMEA?' The AI checks the discount matrix and responds based on regional approval flows. Security-related queries are risky when answered with public data. RAG ensures internal policies guide responses. Examples: Data access controls, employee onboarding/offboarding protocols and acceptable use policy. Use Case: An employee asks how to report a phishing attempt. The AI retrieves and relays the internal incident response protocol and contact escalation path. Final Word In an age where trust, privacy and compliance are business-critical, RAG doesn't just reduce hallucinations—it helps operationalize private knowledge safely across departments. For enterprises betting big on generative AI, grounding outputs in real, governed data isn't optional—it's the foundation of responsible innovation. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Yahoo
04-04-2025
- Business
- Yahoo
Anson Funds names its 3 candidates in board fight at Match Group
By Svea Herbst-Bayliss NEW YORK (Reuters) -Hedge fund Anson Funds blamed Match Group's board for the online dating company's underperformance and named publicly for the first time its three director candidates who all have expertise in the technology sector and have sat on public company boards. In a letter to fellow shareholders, Anson Funds' portfolio manager Sagar Gupta wrote "Match's outdated, insular board with deep interlockings is inadequately qualified to oversee a modern technology company." Reuters obtained a copy of the letter. Anson said it proposed Fumbi Chima, a former senior executive at Adidas , Burberry and Walmart , Laura Lee, who previously worked at Amazon-owned Twitch, Google and consulting firm McKinsey and Kelley Morrell, a former senior managing director at Blackstone, as director candidates. Anson, which owned roughly 0.6% of Match at the end of December according to a regulatory filing, has been pressing the parent of dating sites Tinder, Hinge and OkCupid for over a year to rethink capital allocation, cut costs, and consider a strategic review of its MG Asia business. While Anson's holding is relatively small, lawyers and bankers say the size of an activist's position is less important now as a growing number of investors push for changes at all types of companies. Additionally the investor has raised concerns about Match's governance and pushed for management to refine its corporate strategy. A Match spokesperson was not immediately available for comment. Match's stock price has fallen 10.5% in the last 52 weeks, valuing the company at $7.6 billion. Only three of Match's 10 directors will stand for election this year, creating a flash point for many investors who generally want all board members to be up for election annually. Anson said in its letter that the board played "musical chairs" with the role of chief executive officer, making four changes in the past five years. It also said that it was not eager to fight with the company, noting that "ideally, Match's board will see the need for change itself and work with shareholders to refresh the board." Reuters reported last month that Anson planned to nominate several director candidates. Sign in to access your portfolio
Reuters
04-04-2025
- Business
- Reuters
Anson Funds names its 3 candidates in board fight at Match Group
Summary Companies Anson Funds criticizes Match Group's board for underperformance Anson proposes three tech-savvy director candidates for Match Group Match's stock down 10.5% in past year, valued at $7.6 billion NEW YORK, April 4 (Reuters) - Hedge fund Anson Funds blamed Match Group's (MTCH.O), opens new tab board for the online dating company's underperformance and named publicly for the first time its three director candidates who all have expertise in the technology sector and have sat on public company boards. In a letter to fellow shareholders, Anson Funds' portfolio manager Sagar Gupta wrote "Match's outdated, insular board with deep interlockings is inadequately qualified to oversee a modern technology company." Reuters obtained a copy of the letter. Anson said it proposed Fumbi Chima, a former senior executive at Adidas ( opens new tab, Burberry (BRBY.L), opens new tab and Walmart (WMT.N), opens new tab, Laura Lee, who previously worked at Amazon-owned Twitch, Google and consulting firm McKinsey and Kelley Morrell, a former senior managing director at Blackstone, as director candidates. Anson, which owned roughly 0.6% of Match at the end of December according to a regulatory filing, has been pressing the parent of dating sites Tinder, Hinge and OkCupid for over a year to rethink capital allocation, cut costs, and consider a strategic review of its MG Asia business. While Anson's holding is relatively small, lawyers and bankers say the size of an activist's position is less important now as a growing number of investors push for changes at all types of companies. Additionally the investor has raised concerns about Match's governance and pushed for management to refine its corporate strategy. A Match spokesperson was not immediately available for comment. Match's stock price has fallen 10.5% in the last 52 weeks, valuing the company at $7.6 billion. Only three of Match's 10 directors will stand for election this year, creating a flash point for many investors who generally want all board members to be up for election annually. Anson said in its letter that the board played "musical chairs" with the role of chief executive officer, making four changes in the past five years. It also said that it was not eager to fight with the company, noting that "ideally, Match's board will see the need for change itself and work with shareholders to refresh the board." Reuters reported last month that Anson planned to nominate several director candidates.
Yahoo
11-03-2025
- Business
- Yahoo
Exclusive-Anson Funds eyes proxy fight at Match, plans to nominate directors, sources say
By Svea Herbst-Bayliss NEW YORK (Reuters) - Hedge fund Anson Funds is gearing up for a boardroom fight at Match Group and plans to nominate several directors to the online dating company's 10-member board, two sources familiar with the matter told Reuters. Anson, which owned roughly 0.6% of Match at the end of December according to a regulatory filing, has been pressing the parent of Tinder, Hinge and OkCupid for over a year to rethink capital allocation, cut costs, and consider a strategic review of its MG Asia business, the sources said. Additionally the investor has raised concerns about Match's governance and pushed for management to refine its corporate strategy, said the sources who requested anonymity to speak about the private discussions. Only three of Match's 10 directors will stand for election this year, creating a flash point for many investors who generally want all board members to be up for election annually. Anson also highlighted tight business connections among a handful of Match directors and ties to former owner IAC/Interactive as worrisome and found the quick paced turnover in the executive suite where the company has had four chief executives in five years to be problematic, the sources said. Over the last year, Anson and Match have held a dozen meetings and the company has made some changes, including holding an investor day and agreeing to return capital more aggressively, suggested by activist investors including Anson. A representative for Match was not immediately available for comment and a representative for Anson declined to comment. For Sagar Gupta, who is leading the campaign after he joined Anson as a portfolio manager in 2023 to build the firm's activism practice, the pace of change remains too slow, the sources said. Gupta, who joined the board of U.S. call center software company Five9 in December, has a track record of investing in technology, media and telecommunications. He spearheaded many of those types of investments at activist Legion Partners, where he worked before joining Anson. Match is valued at roughly $8 billion but has shrunk dramatically since the COVID pandemic when it was worth around $40 billion. Its stock price, which closed at $32.03 on Monday, has slipped nearly 2% this year but it has lost 67% in the last three years when the broader S&P 500 stock market index gained 41%. Match's shrinking stock price has made the company vulnerable, prompting at least three activist investors to push for changes within the last year. In early 2024, Elliott Investment Management, one of the world's biggest and most prominent activist investors, unveiled a $1 billion investment. The company named two new directors to the board several weeks later. In July Starboard Value, another activist investor, was urging Match to consider a sale if it could not revitalize the business. Regulatory filings detailing the two activists' holdings at the end of 2024 show Elliott owned a 4.8% stake while Starboard held 5.8% of the company. While Anson's holding is considerably smaller, lawyers and bankers say the size of an activist's position is less important now as a growing number of investors push for changes at all types of companies and corporations are bracing for more expensive and noisy fights with corporate agitators. Sign in to access your portfolio
Reuters
11-03-2025
- Business
- Reuters
Exclusive: Anson Funds eyes proxy fight at Match, plans to nominate directors, sources say
NEW YORK, March 11 (Reuters) - Hedge fund Anson Funds is gearing up for a boardroom fight at Match Group (MTCH.O), opens new tab and plans to nominate several directors to the online dating company's 10-member board, two sources familiar with the matter told Reuters. Anson, which owned roughly 0.6% of Match at the end of December according to a regulatory filing, has been pressing the parent of Tinder, Hinge and OkCupid for over a year to rethink capital allocation, cut costs, and consider a strategic review of its MG Asia business, the sources said. Additionally the investor has raised concerns about Match's governance and pushed for management to refine its corporate strategy, said the sources who requested anonymity to speak about the private discussions. Only three of Match's 10 directors will stand for election this year, creating a flash point for many investors who generally want all board members to be up for election annually. Anson also highlighted tight business connections among a handful of Match directors and ties to former owner IAC/Interactive as worrisome and found the quick paced turnover in the executive suite where the company has had four chief executives in five years to be problematic, the sources said. Over the last year, Anson and Match have held a dozen meetings and the company has made some changes, including holding an investor day and agreeing to return capital more aggressively, suggested by activist investors including Anson. A representative for Match was not immediately available for comment and a representative for Anson declined to comment. For Sagar Gupta, who is leading the campaign after he joined Anson as a portfolio manager in 2023 to build the firm's activism practice, the pace of change remains too slow, the sources said. Gupta, who joined the board of U.S. call center software company Five9 in December, has a track record of investing in technology, media and telecommunications. He spearheaded many of those types of investments at activist Legion Partners, where he worked before joining Anson. Match is valued at roughly $8 billion but has shrunk dramatically since the COVID pandemic when it was worth around $40 billion. Its stock price, which closed at $32.03 on Monday, has slipped nearly 2% this year but it has lost 67% in the last three years when the broader S&P 500 stock market index gained 41%. Match's shrinking stock price has made the company vulnerable, prompting at least three activist investors to push for changes within the last year. In early 2024, Elliott Investment Management, one of the world's biggest and most prominent activist investors, unveiled a $1 billion investment. The company named two new directors to the board several weeks later. In July Starboard Value, another activist investor, was urging Match to consider a sale if it could not revitalize the business. Regulatory filings detailing the two activists' holdings at the end of 2024 show Elliott owned a 4.8% stake while Starboard held 5.8% of the company. While Anson's holding is considerably smaller, lawyers and bankers say the size of an activist's position is less important now as a growing number of investors push for changes at all types of companies and corporations are bracing for more expensive and noisy fights with corporate agitators.
