I lead engineering at KnackLabs. My work spans AI-assisted software delivery and enterprise AI systems, starting with the business problem and the people doing the work.
I’m Ravi Kiran Vemula. At KnackLabs, I lead engineers who work directly with customers. We observe how work happens, question assumptions and uncover decisions that have never been documented.
Customers may have incomplete records, unclear requirements or expectations that AI can solve everything. I help identify the actual problem and choose the right combination of process changes, software and AI.
That experience shapes how I lead: define the expected result, check the important failures and make sure the team can operate what we build. I also lead internal engineering training, create playbooks and mentor the team.
About RaviOur forward-deployed engineers work directly with the customer’s team, at the client’s location when needed, from discovery through production. We choose process changes, software or AI to suit the work.
Illustrative enterprise AI example: A sales team asks for AI to handle quotes. Working with the team reveals varied enquiries, existing customer and pricing systems, and discount rules that only experienced staff know. The solution needs software, AI and human decisions.
For engineering teams: improve software delivery with coding agents.
Understand
We follow a buyer’s quote enquiry from arrival to approval. Quotes wait while staff ask experienced colleagues which discounts apply. We document those decisions and measure the time to an approved quote.
Working directly with the customer
Decide
The AI sales assistant interprets enquiries and drafts replies. Existing software calculates prices. Sales decides exceptions. We reuse the customer and pricing systems they already use, and document missing discount rules.
Process, software and AI each have a role
Build
Our engineers build the quote workflow with the sales team and test it on sample enquiries. We use Symphony Forge, our software factory process, for implementation, checks and releases.
Built using Symphony Forge
Run
Gantry manages the AI sales assistant’s access, approvals and activity record. When a buyer requests a volume discount outside agreed terms, sales reviews the draft, approves the discount and records the conditions under which it applies.
Operated through Gantry
Improve
When a new enquiry matches the recorded conditions, the AI sales assistant suggests the discount for sales to approve. We review time to an approved quote and repeated corrections with the customer, then improve the guidance, data and software.
Sales still approves exceptions
Built at KnackLabs · Open source
A runtime for AI employees. Teams define their responsibilities, control access to tools and data, approve actions and review activity.
Built at KnackLabs · Open source
The software factory our engineers use to build with coding agents. Agreed requirements, implementation, checks and human review stay connected through release.
Built · Engineering skills and fresher hiring
An AI-taught bootcamp for freshers, from system design to debugging and production readiness. Students build projects and evidence that hiring teams can assess.
Start with what is slowing the work down. We can work out what needs to change.
Improve requirements, implementation, review and testing. Measure delivery time, review effort and rework.
Understand the business process, prepare the data and systems, then build a solution the team can operate.
I speak about agent systems and software delivery, and host LangChain Connect India meetups.
Notes on enterprise AI, coding agents and software delivery.
All writingTell me about the problem and the people it affects.
