Ramanchi Rishi Sai Teja@rishisaiteja
1
Events attended
1
Submissions made
India
1 year of experience
About me
I am a Computer Science Engineering graduate with a strong focus on applied artificial intelligence, data engineering, and scalable software systems. My work centers on building production-ready AI solutions, robust data pipelines, and analytics-driven applications that translate complex data into measurable business impact. I have hands-on experience designing and implementing end-to-end machine learning workflows, including sentiment analysis, context-aware NLP systems, time-series forecasting using ARIMA, SARIMA, and SARIMAX models, and LLM evaluation frameworks. I have worked extensively with Microsoft Fabric, Power BI, Azure AI services, and modern cloud-based data platforms to build interactive dashboards, ETL pipelines, and automated AI-driven workflows. My technical stack includes Python, Java, C, C++, JavaScript, SQL, React.js, and enterprise data tools across the Microsoft ecosystem. I am comfortable working across the complete development lifecycle from problem definition and architecture design to deployment, optimization, and stakeholder delivery. I value clean system design, maintainable architecture, and practical AI implementations that solve real-world business problems rather than theoretical experiments. I completed my Bachelor of Engineering in Computer Science and Engineering from Matrusri Engineering College and am currently seeking opportunities as an Applied AI Engineer, Data Engineer, or Machine Learning Engineer in technically demanding environments that value ownership, depth, and measurable impact.
🤓 Latest Submissions
Sentinel — Live Threat Hunting Agent
Sentinel is an autonomous defensive-security agent that gives security teams live coverage of a surface their internal tools weren't built to monitor: the open web. Leaked credentials, impersonation domains, phishing kits, and brand-abuse signals don't surface in your SIEM — they surface across paste sites, forums, lookalike domains, and search results. Sentinel hunts them autonomously and returns prioritized, structured findings a security team can act on. How it works. Given a company name and primary domain, a LangChain-powered agent runs a streaming, multi-stage loop: it plans targeted defensive search queries, runs them through the Bright Data SERP API, triages which results are worth investigating, fetches each page through the Bright Data Web Unlocker (past bot detection, CAPTCHAs, and geo-blocks), and reasons over the extracted content to assess whether it represents a real exposure signal. Findings are scored by severity, synthesized into an executive risk report with recommended actions, and persisted to Cognee memory so subsequent scans flag only NEW exposures since the last run — the difference between alert fatigue and actionable intelligence. Why this matters. Most builders reach for sales tools; very few build for the security team. Yet open-web threat surface is exactly the gap Bright Data's infrastructure is uniquely positioned to close, because it bypasses the bot defenses that block naive scrapers. Sentinel demonstrates this gap in a workflow enterprise security teams would genuinely depend on. Defensive by design. Sentinel only surfaces information that is already public, never reproduces sensitive values it encounters, and is scoped to organizations you own or are authorized to monitor. A Streamlit UI streams the agent's reasoning live — plan, search, fetch, analyze — so you can watch it think before the report renders.
31 May 2026
