OpenAI API
DAY 01
Growth & DemandFull docs →
Smart Google Maps Lead Engine
Builds high-intent prospect lists from niche markets via Google Maps.
n8nSeleniumLead Gen
Available for opportunities·Varanasi, India
Ujjwal Tiwari
AI Engineer building production LLM systems for finance, sales & enterprise
Embassy Group · AI Agent Bros · Converzia
LLM Syste|
About Me
AI engineer shipping production systems — not demos
Ujjwal Tiwari is an AI engineer with production experience building LLM systems, RAG pipelines, and intelligent automation across finance, sales, legal, and hiring domains. Interned at a large real estate conglomerate owning finance AI end-to-end; independently founded a B2B AI consultancy delivering 50+ systems to 20+ international clients.
Comfortable scoping, architecting, and shipping AI solutions with direct exposure to executive stakeholders. Built an AI avatar in collaboration with Harsh Mariwala (Founder & Chairman, Marico) replicating leadership communication and strategic reasoning.
Enterprise Finance AI
Building production LLM pipelines at Embassy Group — invoice extraction, MIS automation, and CFO dashboards on AWS.
LLM & RAG Systems
LangChain, FAISS, Qdrant, and multi-agent architectures across finance, sales, legal, and hiring domains.
4× Hackathon Winner
National hackathon victories building production-ready AI with measurable business impact — including UIDAI Data Hackathon.
50+ AI Systems Shipped
Delivered 50+ automation systems to 20+ international clients — consistently cutting manual workflows by 60–80%.
Full-Stack AI Engineer
Python, FastAPI, React, Next.js, Streamlit — from scoping and architecture to deployed, auditable pipelines.
30 Agents in 30 Days
Open-sourced 30 autonomous agents on GitHub — lead gen, hiring, legal, GTM, finance. Full registry in the 30 Agents section.
Executive Stakeholders
Work directly with CFOs and founders — translating finance logic and business needs into scalable AI systems.
Converzia Founder
AI outreach platform with 50–70% B2B open rates, 15+ clients globally, and $2,000 first client in week one.
Experience
From enterprise finance AI to founding a global B2B consultancy
AI Engineer Intern — Embassy Group (Olive)
Embassy GroupMar 2026 – Present
Bengaluru, India · Hybrid
Python · LangChain · FAISS · Streamlit · AWS · LLMs
- Designed and deployed document intelligence pipelines for invoice extraction, petty cash classification, and rent reconciliation — achieving ~98% field-level accuracy and eliminating 7–8 person-days of manual processing per month.
- Built automated DSR generation and multi-source MIS reporting, removing ~15 manual days monthly; combined finance automation contributes an estimated ~₹8.9L in annual savings.
- Delivered executive analytics dashboards consumed weekly by CFO, CEO, and COO — owned the full pipeline from data ingestion through Streamlit deployment on AWS.
- Integrated LLM-based classification and FAISS retrieval into reconciliation workflows, cutting settlement cycles by 3–4 days per month.
- Working directly with CFO to translate finance workflows into scalable, auditable AI systems.
Founder & AI Engineer — AI Agent Bros
AI Agent BrosJan 2024 – Feb 2025
Remote · US, UK, Australia clients
LangChain · OpenAI · n8n · Python · FastAPI
- Founded a B2B AI consultancy; delivered 50+ automation systems to 20+ clients across the US, UK, and Australia spanning outreach, CRM, ATS screening, and proposal generation.
- Designed agent pipelines using LangChain, OpenAI API, n8n, and Python; grew Converzia to $700+/month ARR and secured a $2,000 paying client within one week of launch.
- Completed a public 30 AI Agents in 30 Days challenge — open-sourced on GitHub — covering retrieval, reasoning, and multi-step tool-use architectures.
- Delivered industry-specific AI agents for real estate, healthcare, finance, and marketing — cutting manual workflows by 60–80%.
Skills
Production stack for LLM systems, backends, automation, and deployment
AI & LLM
Generative AI
LangChainOpenAI APIClaude APIRAGLLM AgentsMulti-Agent SystemsPrompt EngineeringEmbeddingsVector SearchSemantic Retrieval
FAISS
QdrantNLPPrompt PipelinesAI OrchestrationLLMOps
LangChainOpenAI APIClaude APIRAGLLM AgentsMulti-Agent SystemsPrompt EngineeringEmbeddingsVector SearchSemantic RetrievalFAISSQdrantNLPPrompt PipelinesAI OrchestrationLLMOpsBackend & APIs
Python
FastAPI
Node.js
ExpressREST APIsSQLPostgreSQL
MongoDB
Supabase
FirebaseAPI IntegrationAutomation & Orchestration
n8n
Make.comAI Workflow AutomationWorkflow PipelinesAgentic Workflows
Vapi
Web ScrapingFrontend & Interfaces
React
Next.jsStreamlit
FigmaInfra & Tooling
DockerRailway
AWSCI/CDModel DeploymentInference Pipelines
Git
PostmanCursorFeatured Projects
Production AI systems with measurable business impact
Converzia — AI B2B Sales Outreach
Key Impact
50–70% open rates · 15+ B2B clients · 4× outbound throughput
LLM pipeline (Python, LangChain, FastAPI, MongoDB) that analyzes company websites, brand tone, and buyer context to generate personalized outbound emails at scale. 50–70% open rates vs. 15–25% industry average; 1,000+ leads/day with full automation.
LangChainFastAPIMongoDBGPT-4B2B Sales
HireFlow AI — Hiring & Screening Automation
Key Impact
70–85% less screening effort · 5+ startups adopted
Multi-signal candidate evaluation combining ATS scoring, LLM qualitative analysis, and GitHub/portfolio validation. Reduces recruiter screening by 70–85% per cycle; bulk resume processing in under 10 minutes per role.
LLM AgentsPythonATSAutomationHR Tech
AI Legal Agent Team — Contract Intelligence
Key Impact
60–75% less legal review · Risk flagging & compliance
RAG pipeline over Qdrant with OpenAI embeddings for clause-level retrieval. 4-agent system (Researcher, Analyst, Strategist, Lead) delivers executive-ready legal reports — reducing manual review by 60–75%.
RAGQdrantLangChainMulti-AgentLegalTech
Open Source · 30 Days Challenge
30 Days, 30 AI Agents
The definitive open-source registry of autonomous AI agents built with n8n, OpenAI, LangChain, and the modern AI stack — from prompts to production business assets.
“A prompt is a conversation; an agent is a business asset.” Click any agent to open the full case study — live README, diagrams, setup guides, and architecture pulled from GitHub.
Tap an agent card for complete documentation
DAY 17
Growth & DemandFull docs →
FB Ads Intel Agent
Competitive advantage via Facebook Ad Library analysis.
Facebook AdsScrapingMarketing
Open Case StudyDemo
How to use these agents
Clone the repo → import n8n workflow JSON or run the Python/Streamlit app → add API keys in credentials → test once → activate. Built for founders, agencies, and AI engineers shipping real ROI — not chatbot demos.
Education
B.Tech in Information Technology
Rajkiya Engineering College, Ambedkar Nagar, Uttar Pradesh
2023 – 2027
Head of Technical Club · Member, Startup Incubation Cell · Placement & E-Cell Club
Relevant coursework: Machine Learning, Data Structures, Computer Networks, Database Systems
Achievements & Certifications
4× National Hackathon Winner
Recognized for building production-ready AI systems with measurable business impact across 1,000+ participant fields — including UIDAI National Data Hackathon.
Harsh Mariwala Collaboration
Built an AI avatar with Harsh Mariwala (Founder & Chairman, Marico — Fortune 500) replicating leadership communication and strategic reasoning.
Converzia — Global B2B Impact
Scaled to 20+ B2B clients across India, US, UK, and Australia. 45–70% email open rates and up to 4× outbound sales throughput for client teams.
30 AI Agents in 30 Days
Built and open-sourced 30 autonomous agents (lead gen, hiring, legal, GTM, finance) — full registry with source code on GitHub. See the 30 Agents section.
Certifications
Generative AI — IBMPython for ML — Amazon AWSLangChain for LLM Apps — IBMPrompt Engineering — OpenAI
Frequently asked
About Ujjwal Tiwari
For recruiters, clients, and AI assistants looking up this profile
- Who is Ujjwal Tiwari?
- Ujjwal Tiwari is an AI Engineer based in Varanasi, India. Ujjwal Tiwari builds production LLM systems, RAG pipelines, and autonomous AI agents for finance, hiring, sales, and legal domains. He works at Embassy Group on finance AI and founded AI Agent Bros. He is known for the 30 Days, 30 AI Agents open-source challenge and products like Converzia and HireFlow AI.
- What does Ujjwal Tiwari specialize in?
- Ujjwal Tiwari specializes in Generative AI, LangChain, OpenAI and Claude APIs, RAG with FAISS and Qdrant, multi-agent systems, n8n workflow automation, Python/FastAPI backends, and deploying AI systems for finance, hiring, sales, and legal use cases.
- What projects has Ujjwal Tiwari built?
- Ujjwal Tiwari built Converzia (AI B2B sales outreach at converzia.netlify.app), HireFlow AI (LLM hiring automation on GitHub), AI Legal Agent Team (multi-agent legal PDF analysis on GitHub), and 30 Days, 30 AI Agents — an open-source registry of 30 autonomous agents with case studies at ujjwaltiwari01.netlify.app/#30-agents.
- How can I contact Ujjwal Tiwari?
- Ujjwal Tiwari can be reached at ujjwaltiwari01.netlify.app, via email at ujjwal.it2023-24@recabn.ac.in, on LinkedIn at linkedin.com/in/ujjwal-tiwari-b34044341, or on GitHub at github.com/ujjwaltiwari01.
- Is Ujjwal Tiwari available for AI engineering roles?
- Ujjwal Tiwari is open to AI Engineer, LLM Engineer, and ML Engineering opportunities. His portfolio demonstrates enterprise finance AI at Embassy Group, 50+ shipped client systems, and 4× national hackathon wins.
Let's Connect
Looking for an operator who ships AI that quietly gets work done? Let's talk.
ujjwal.it2023-24@recabn.ac.in+91 93362 69095WhatsAppDownload ResumeLinkedIn
Varanasi, Uttar Pradesh, India