Design and implement intelligent exception resolution capabilities by integrating LLMs like Claude and OpenAI into a financial reconciliation platform. Develop robust API integrations and human-in-the-loop workflows to ensure financial accuracy and auditability.
AI/ML Engineer
Company: CODVO.AI
Department: Engineering
Location: Work From Home
Employment Type: Full-Time
Experience Level: Mid-Level (3+ years)
Shift Timing: 9:30 AM - 6:30 PM IST
Role Overview
We're seeking an experienced AI/ML Engineer to design and implement intelligent exception resolution capabilities within our financial reconciliation platform. You'll integrate advanced language models (Claude/OpenAI) into our core pipeline, crafting sophisticated prompts that transform complex reconciliation exceptions into clear, auditable resolution suggestions. This role is critical to delivering AI-powered insights that help financial teams resolve discrepancies faster and with greater confidence.
What You'll Do
Design and optimize prompts that accurately interpret flagged reconciliation exceptions and generate structured, actionable resolution suggestions with confidence scoring
Integrate LLM APIs (Anthropic Claude, OpenAI) into the exception resolution layer using Python, the Anthropic SDK, and async API patterns for production-grade reliability
Develop human-in-the-loop workflows that balance automation with human oversight, ensuring financial accuracy and auditability at every step
Build robust API integration code that handles edge cases, rate limiting, token management, and graceful error recovery in a financial context
Collaborate with backend developers to embed AI outputs cleanly into the reconciliation pipeline and ensure seamless data flow into final reports
Implement LLM evaluation and benchmarking frameworks to measure prompt effectiveness, accuracy, and consistency across diverse exception types
Document prompt strategies and model behavior to enable knowledge sharing and continuous improvement across the engineering team
Monitor and iterate on model performance, refining prompts and integration logic based on real-world exception data and user feedback
What We're Looking For
Required Qualifications:
3+ years of professional software engineering experience, with at least 2+ years working directly with LLM APIs (Anthropic, OpenAI, or similar)
Strong proficiency in Python, including hands-on experience with the Anthropic SDK and async API calls
Demonstrated expertise in prompt engineering, particularly for structured outputs and domain-specific tasks
Solid understanding of confidence scoring, uncertainty quantification, and human-in-the-loop AI workflows
Experience writing clean, maintainable API integration code with attention to edge cases and error handling
Ability to work independently and communicate effectively with cross-functional teams
Familiarity with financial systems, reconciliation processes, or similar regulated domains is a plus
Technical Skills:
Backend development and ETL middleware concepts
Agent orchestration and agentic AI patterns
LLM evaluation and benchmarking methodologies
Version control (Git) and collaborative development practices
Nice to Have
Experience with Retrieval-Augmented Generation (RAG) or domain-specific knowledge grounding
Familiarity with financial reconciliation, accounting systems, or fintech platforms
Experience with prompt versioning and A/B testing frameworks
Knowledge of structured output formats (JSON, XML) and validation
Exposure to observability and monitoring tools for LLM applications
Background in machine learning model evaluation and metrics
What We Offer
Flexible work environment: 100% remote with a collaborative, inclusive team culture
Meaningful impact: Work on cutting-edge AI solutions that directly improve financial operations for our clients
Professional growth: Opportunities to deepen expertise in LLMs, agentic AI, and financial technology
Competitive compensation: Commensurate with experience and market rates
Collaborative culture: Work alongside talented engineers and product teams who value innovation and continuous learning
Billable engagement: Full-time, stable project with mid-to-late stage engagement timeline
About CODVO.AI
CODVO.AI is a forward-thinking technology company specializing in AI-driven financial solutions. We empower organizations to automate and optimize their reconciliation processes through intelligent, trustworthy AI. Our platform combines advanced language models with domain expertise to deliver accurate, auditable insights that save time and reduce errors. We're committed to building inclusive, innovative products that solve real-world problems in the financial services space.
Note- Please apply via our official careers portal only, as applications sent directly to executives may not be considered.
“I was the first applicant for a remote marketing position that got listed on the company website the same day I applied. Had an interview within 48 hours!”