Mudassir A.

I’m a Principal AI Engineer who has faced the ups, downs, and worst problems of shipping AI agents to production, and the number one thing you need to start with is "DATA". (Almost 5 years building only AI applications)

𝐑𝐞𝐜𝐞𝐧𝐭 𝐰𝐨𝐫𝐤:

• 𝐀𝐧 𝐀𝐈 𝐜𝐨-𝐟𝐨𝐮𝐧𝐝𝐞𝐫 𝐟𝐨𝐫 𝐬𝐭𝐚𝐫𝐭𝐮𝐩𝐬 𝐚𝐧𝐝 𝐒𝐌𝐁𝐬: specialist agents for planning, fundraising and operations, backed by long-term business memory, graph + vector retrieval and proactive workflows. It connects financial data through Xero, Plaid and QuickBooks, alongside Gmail, Calendar and meeting data, continuously building business context, tracking progress and deciding what needs attention next, while keeping the interface as simple as WhatsApp and voice calling

• 𝐕𝐨𝐢𝐜𝐞 𝐀𝐈 𝐟𝐨𝐫 𝐢𝐧𝐭𝐞𝐫𝐚𝐜𝐭𝐢𝐯𝐞 𝐰𝐨𝐫𝐤𝐩𝐥𝐚𝐜𝐞 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠: real-time AI role-play environments built with Agora/WebRTC, Whisper + Deepgram speech recognition and ElevenLabs voice, including turn-taking, conversational memory and scenario-driven learner progression. Organisations can provide their own material and generate customised training scenarios and AI characters in English, Dutch, German, and Spanish, delivered via the web or embedded in LMS platforms through iFrame/LTI integrations.

• 𝐀𝐈 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 𝐨𝐯𝐞𝐫 𝐚 𝐠𝐨𝐯𝐞𝐫𝐧𝐦𝐞𝐧𝐭 𝐭𝐫𝐚𝐢𝐧𝐢𝐧𝐠 𝐚𝐧𝐝 𝐪𝐮𝐚𝐥𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐫𝐞𝐠𝐢𝐬𝐭𝐞𝐫: production ETL across 150,000+ current and historical records, with MongoDB, Azure AI Search, embeddings and a custom MCP text-to-NoSQL layer combining structured queries with semantic retrieval. A plan-first agent decomposes complex questions into parallel/sequential retrieval steps, tracks source changes and supersession relationships, and compiles complete, evidence-referenced reports and impact analysis.

• 𝐑𝐞𝐠𝐮𝐥𝐚𝐭𝐨𝐫𝐲 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 𝐨𝐯𝐞𝐫 𝐔𝐒 𝐞𝐧𝐞𝐫𝐠𝐲 𝐟𝐢𝐥𝐢𝐧𝐠𝐬: ingestion and typed extraction over large regulatory dockets, with graph relationships connecting filings, proceedings and entities across an evolving taxonomic hierarchy. Validation gates, structured extraction contracts and evidence checks sit before anything reaches a count, timeline or generated report.

• 𝐀𝐳𝐮𝐫𝐞-𝐨𝐧𝐥𝐲 𝐢𝐧𝐭𝐞𝐫𝐧𝐚𝐥 𝐀𝐈 𝐚𝐬𝐬𝐢𝐬𝐭𝐚𝐧𝐭 𝐟𝐨𝐫 𝐚 𝐠𝐥𝐨𝐛𝐚𝐥 𝐛𝐞𝐚𝐮𝐭𝐲 𝐫𝐞𝐭𝐚𝐢𝐥𝐞𝐫: serving an organisation of 10,000+ employees with hybrid retrieval across policies, procedures and internal knowledge. Department-level access control, grounded answers and the full AI/data stack run inside the client's own Azure tenancy under GDPR and CCPA constraints.