WHY I’M A FIT FOR CONFIDO
I’m Anudeep, a product-minded software engineer who takes unclear business problems and turns them into dependable products, automations, and agentic workflows.
At , I build multi-tenant systems for portfolio businesses and internal teams. My work includes executive reporting agents, privacy-aware request routing, developer operations, APIs, permissions, and full-stack product development.
I use Claude Code and Codex throughout development. Claude Code is my primary tool for implementation, debugging, and testing. I use Codex for codebase exploration, documentation, and rapid prototyping, and I use both when reviewing security.
Confido needs an engineer who can understand an operation, decide where AI is useful, connect the required systems, and ship a working product quickly.
That closely matches my current work. I have built agents that retrieve data through APIs and MCP connectors, make bounded decisions, manage structured state, route exceptions, validate their own output, and ask for human approval when judgment matters.
I also understand the less visible parts of these products. I work with authentication, row-level security, role-based routes, masked customer information, multi-tenant data, Microsoft 365 permissions, deployment, and support after release.
I have shipped multi-step workflows that retrieve data, use tools, manage state, follow routing rules, validate outputs, retry failures, and involve people when judgment is required.
I built Evalutron's first working version in two days and Portfolio Intelligence's production MVP in five. I start with the smallest useful system, put it in front of real users, and improve it from evidence.
I work from the operational problem through interface design, APIs, backend services, permissions, automation, deployment, documentation, and ongoing support.
I have built role-based, multi-tenant products with row-level security, permission-controlled routes, masked customer data, validation checks, audit trails, and human approval steps.
USE CASES
How I use AI, automation, product thinking, and software engineering to solve real problems.
AI OPERATIONS · EXECUTIVE REPORTING
A multi-step reporting agent that turns operational and financial data across 8+ properties into validated weekly executive updates.
Approximately 16 hours of manual reporting saved each week
AI OPERATIONS · REQUEST ROUTING
A privacy-aware routing agent that classifies website inquiries and assigns each request using live workload and permission data.
Productivity improved ~60%; response times cut to 12–24 hours
DEVELOPER TOOLS · AGENTIC WORKFLOWS
An AI-assisted operational layer that connects sprint plans with repository activity to automate documentation, ownership, and release communication.
Recurring sprint meetings reduced from five to two per week
EDUCATION · AI ASSESSMENT
A human-reviewed GPT-4 assessment agent that evaluates university submissions against controlled rubrics and valid solution paths.
A working v1 built in two days for a 210-student course
DEVELOPER TOOL · PROMPT ENGINEERING
A private prompt-engineering workspace that semantically compresses, structures, measures, and saves reusable prompts.
Typical prompt token usage reduced by 20–40%
PERSONAL AI · SOCRATIC LEARNING
A scheduled, context-aware Socratic learning agent that adapts its questions, tracks understanding, and saves progress to Apple Notes.
80% of a system-design curriculum completed
PERSONAL AI · HEALTH TRACKING
A scheduled Codex agent that uses Apple Watch data to surface longitudinal fitness and recovery patterns without making medical claims.
Persistent, context-aware health and fitness reflection
PRODUCT ENGINEERING · SUSTAINABILITY
A website sustainability product that translates carbon-emission and green-hosting APIs into clear audits and locally saved history.
More than 1,000 views in one week
THE FIT
I have already worked in the way this role describes. I identify an operational problem, learn how the people and systems around it work, build the first useful version quickly, and stay responsible for the result after it ships.
My projects are not AI demonstrations. Portfolio Intelligence supports weekly executive decisions. Business Query Router assigns real customer inquiries. Developer Operations Copilot keeps an engineering team aligned. Evalutron helped a university grading team recover from a large backlog.
I would bring that same balance of engineering depth, product judgment, security awareness, and urgency to Confido.