Chekit Sharma profile photo 1 on Tal

Chekit Sharma

SDE 1 @ Ola | Former AI/ML Intern @ Ola

Bengaluru

< 1 year of work experience

Ola
Aug 20262 mosPresent

Bengaluru, Karnataka, India

Ola Electric
Feb 20267 mosAug 2026

India

- Built and maintained LLM-based voice agents for sales outreach, lead qualification, and service follow-up use cases — covering prompt design, tool configuration, and knowledge-base integration. - Cut AI calling costs by optimizing prompts and LLM templates for token efficiency, and tuned text-to-speech settings to improve call quality and naturalness. - Designed a scheduler-driven agent that analyzes recent call transcripts, surfaces prompt-related issues automatically, and publishes an optimized prompt version for review — reducing manual prompt-debugging effort. - Led migration of production voice agents from a third-party provider (ElevenLabs) to an in-house platform on Vertex AI, integrating a retrieval-based knowledge service to power agent responses at a lower per-call cost. - Built an LLM-based backend wrapper that auto-generates production-ready calling agents from minimal input, with a UI enabling HR, Service, and Marketing teams to launch their own campaigns — plus a cost-estimation feature and a live spend-tracking dashboard integrated with cloud billing data. - Owned the ticket-creation workflow for Service Safai (EV service operations) end-to-end — models, APIs, validation, and tests — and migrated the serviceType field from Ticket to JobCard across schema, DTOs, and OTP logic. Standardized API error responses with OpenAPI documentation.
Ola Electric
Feb 20267 mosAug 2026

Set up, configured, and maintained LLM-based voice agents for business use cases (sales outreach, lead qualification, service follow-ups), covering prompt design, tool configuration, knowledge-base mapping, and production issue resolution. Reduced AI calling costs by systematically optimizing prompts and LLM templates for token efficiency, and tuned text-to-speech settings for more natural outbound calls. Built a scheduler-driven agent that periodically analyzes recent call transcripts to surface prompt-related issues and automatically publishes an optimized prompt version on the same agent for review. Supported migrating production voice agents from a third-party provider (ElevenLabs) to an in-house platform on Vertex AI, and integrated a retrieval-based knowledge service to power agent responses, at a lower per-call cost. On the call-analytics backend, built an LLM-based wrapper that generates production-ready agents from minimal input, with a UI enabling teams (HR, Service, Marketing) to launch calling campaigns; added a cost-estimation feature and a separate cost-tracking tool integrating cloud billing data with a live spend dashboard. For Service Safai (EV service operations), built the ticket-creation workflow (models, APIs, validation, tests), migrated the serviceType field from Ticket to JobCard (schema, DTO, and OTP changes), and standardized API error responses with OpenAPI documentation.
GitHub
@Truelove27

Chekit’s AI Stack

claude code
claude code
codex
codex
cursor
cursor

Education

Atria Institute of technology

Bachelor of Engineering, Information Science/Studies

Atria Institute of technology

20222022
kendriya vidyalaya

High School Diploma

kendriya vidyalaya

Join Chekit and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Chekit Sharma on Tal

Headline

SDE 1 @ Ola | Former AI/ML Intern @ Ola

About

As an Artificial Intelligence Engineer at Ola Electric, I contribute to the development and maintenance of LLM-based voice agents tailored for sales outreach, lead qualification, and service follow-ups. My role includes optimizing prompts and LLM templates for token efficiency, designing scheduler-driven agents that streamline debugging, and supporting the migration of voice agents to an in-house platform on Vertex AI. With a Bachelor of Engineering in Information Science from Atria Institute of Technology, my work centers on aligning AI technologies with business objectives. Dedicated to leveraging prompt engineering and internal tooling, I aim to enhance operational efficiency, reduce costs, and improve the usability of AI-driven solutions in enterprise settings.

Location

Bengaluru

AI stack

claude code

codex

cursor

Work experience

SDE 1 at Ola, Aug 2026 - Present · 2 mos, Bengaluru, Karnataka, India

Artificial Intelligence Engineer at Ola Electric, Feb 2026 - Aug 2026 · 7 mos, India, - Built and maintained LLM-based voice agents for sales outreach, lead qualification, and service follow-up use cases — covering prompt design, tool configuration, and knowledge-base integration. - Cut AI calling costs by optimizing prompts and LLM templates for token efficiency, and tuned text-to-speech settings to improve call quality and naturalness. - Designed a scheduler-driven agent that analyzes recent call transcripts, surfaces prompt-related issues automatically, and publishes an optimized prompt version for review — reducing manual prompt-debugging effort. - Led migration of production voice agents from a third-party provider (ElevenLabs) to an in-house platform on Vertex AI, integrating a retrieval-based knowledge service to power agent responses at a lower per-call cost. - Built an LLM-based backend wrapper that auto-generates production-ready calling agents from minimal input, with a UI enabling HR, Service, and Marketing teams to launch their own campaigns — plus a cost-estimation feature and a live spend-tracking dashboard integrated with cloud billing data. - Owned the ticket-creation workflow for Service Safai (EV service operations) end-to-end — models, APIs, validation, and tests — and migrated the serviceType field from Ticket to JobCard across schema, DTOs, and OTP logic. Standardized API error responses with OpenAPI documentation.

Intern at Ola Electric, Feb 2026 - Aug 2026 · 7 mos, Set up, configured, and maintained LLM-based voice agents for business use cases (sales outreach, lead qualification, service follow-ups), covering prompt design, tool configuration, knowledge-base mapping, and production issue resolution. Reduced AI calling costs by systematically optimizing prompts and LLM templates for token efficiency, and tuned text-to-speech settings for more natural outbound calls. Built a scheduler-driven agent that periodically analyzes recent call transcripts to surface prompt-related issues and automatically publishes an optimized prompt version on the same agent for review. Supported migrating production voice agents from a third-party provider (ElevenLabs) to an in-house platform on Vertex AI, and integrated a retrieval-based knowledge service to power agent responses, at a lower per-call cost. On the call-analytics backend, built an LLM-based wrapper that generates production-ready agents from minimal input, with a UI enabling teams (HR, Service, Marketing) to launch calling campaigns; added a cost-estimation feature and a separate cost-tracking tool integrating cloud billing data with a live spend dashboard. For Service Safai (EV service operations), built the ticket-creation workflow (models, APIs, validation, tests), migrated the serviceType field from Ticket to JobCard (schema, DTO, and OTP changes), and standardized API error responses with OpenAPI documentation.

Education

Bachelor of Engineering, Information Science/Studies, Atria Institute of technology

High School Diploma, kendriya vidyalaya

Public links

github: https://github.com/Truelove27