Paving the Road to Production: Governed Enterprise AI Agents — Harish Gaggar, Credit Karma
Read the talk
Paving the Road to Production: Governed Enterprise AI Agents
Harish Gaggar walks through an enterprise agent platform at Intuit Credit Karma, from registration and authenticated tool access to checkpoints, evaluation records, sensitive-input filtering and context optimization.
From a talk by Harish Gaggar
At a glance
Ideas worth remembering
A paved path should create supporting resources as well as code: workload identity, memory storage, tracking tables and secret management need to be ready when the agent deploys.
The MCP gateway centralizes authentication and routing; reviewed MCP tools constrain the operations and data available behind it.
Policy branches should distinguish execution from auditing: a rejected sensitive action still needs a record, even though it does not run.
Threads preserve conversational continuity, runs identify individual executions, and checkpoints save workflow state at graph nodes.
The demonstrated optimizer reduced one query from 24,000 to 2,400 tokens. That establishes a token reduction for the example, without establishing preserved answer quality or business ROI.
The missing foundations around a working agent
An agent can answer a question locally while leaving its most consequential production questions unresolved. Who authenticated the tool call? Which operations may the tool perform? Can a follow-up recover the previous conversation? Did an API key enter the model prompt—or get copied into a trace? Harish Gaggar, a senior staff software engineer responsible for AI analytics infrastructure and services at Intuit Credit Karma, opens with these gaps between an experiment and an enterprise deployment.
The proposed response is a collection of supporting frameworks, rather than a “magical agent.” LangGraph supplies task-specific nodes and workflow state. An MCP gateway authenticates requests and routes them to custom MCP servers, whose tools perform the operations. Memory, tracking and evaluation, sensitive-data filtering and context optimization handle concerns that a functioning prompt does not resolve by itself.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Registration creates the repository and its production resources
The paved path starts before implementation. A developer registers the agent’s use case and receives an agent ID. In an internal developer portal, they supply that ID, the intended use case, a framework such as LangGraph or LangChain, a language such as Python or TypeScript, and the skills or prompts they want to use. Those choices become inputs to repository generation.
The generated repository includes the agent framework together with memory, tracking and evaluation, context optimization and PII filtering. This addresses a predictable inconsistency: individual developers may implement only the pieces their coding assistant suggests, or concentrate on one or two frameworks. Supplying the surrounding components at creation time gives agents a consistent starting point and makes their results easier to inspect centrally.
Development then moves through a concrete release sequence. The developer clones the repository, builds a Docker image and tests query inputs and outputs with the supporting storage. Once the agent works locally, a pull request requires two approvals. The release process registers the agent centrally and provides a Langfuse tracing URL; CI/CD builds the production container and scans it for vulnerabilities. After security approval, the image enters the registry and an internal orchestrator deploys it into a Kubernetes namespace. Scanning is a release control, not proof that an image contains no vulnerability.
The portal also provisions the workload identity, memory database, tracking and evaluation tables, and secret management. That provisioning matters at deployment: the agent should already have its identity binding and know which resources to use. The orchestrator can deploy an agent with its supporting resources prepared, rather than leave each team to connect identity, storage and tracing afterward.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Reviewed tools behind a centralized MCP gateway
Custom MCP servers provide a place to control what tools expose. Gaggar recommends building them internally for an air-gapped environment and describes using a FastMCP wrapper to define the tools. Each tool undergoes internal security review before production use. Filters, limits and excluded fields can then constrain what the tool returns, including information that should never reach the model. The decision is about the operations and data exposed by a tool, not merely whether its server runs inside the company.
Both users and service accounts must authenticate through the MCP gateway. At the first query, a coding assistant or agent sends an authentication request. The gateway saves session state in Redis and returns a JWT access token to the client. This establishes a session before the client requests a tool operation.
For the tool request, the client supplies the JWT as a bearer token and addresses a namespace identifying the intended MCP server—for example, an Artifactory or GitHub namespace. The gateway uses that namespace to select the server and attaches the user token to the forwarded request. The MCP server calls the downstream service, handles the described token exchange and returns data within the user’s authorized scope. It then packages the response for the gateway to return to the client.
Authentication and routing therefore happen on the same required path: client, gateway, MCP server, downstream endpoint. The intended design prohibits clients from bypassing the gateway to call those endpoints directly. A central entry point makes authentication consistent across local coding tools and deployed agents, while the reviewed server tools control the operations available behind it.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Separate operational logs, request traces and aggregate metrics
A centralized gateway also creates a useful place to observe failures. Its monitoring separates three kinds of evidence, each answering a different operational question:
- Operation logs: a store such as ELK, Loki or CloudWatch records events that help investigate failures such as an unsuccessful token refresh.
- Request traces: OpenTelemetry collects and processes trace data for Tempo. A trace helps identify which tool was slow, whether a request originated in an agent or an IDE, and where a particular call failed.
- Aggregate metrics: Prometheus stores measurements such as a tool’s P95 latency and changes in authentication rejection rates. These reveal patterns across calls rather than the history of one request.
Grafana supplies the visualization layer for the traces and metrics. The distinction is practical: an authentication rejection spike tells an operator that something changed; a trace and the corresponding operation logs help explain which request failed and why. These gateway records describe access and execution behavior. Later, the agent’s own tracking tables address questions about the answer, model use and cost.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Classify the operation before allowing it to execute
LangGraph provides loops, branches and state management for the agent workflow. The first stages interpret the question and form a plan, then classify the requested operation: reading information, modifying it, deleting it or performing another sensitive action. That classification changes how the plan proceeds. The plan itself is a subgraph containing the substantive work.
Evidence gathering uses the same authenticated tool path described earlier. The workflow calls the MCP gateway, reaches the appropriate MCP server and retrieves data through a tool. After validating the findings, it checks policy. Read-only operations can proceed to execution and auditing; write or sensitive operations wait for human approval. Approval permits execution. Rejection still produces an audit record, so the record includes decisions that stopped an action as well as actions that ran.
Where does human approval change the outcome? The branch below separates permission to execute from the obligation to record the decision. Both approval and rejection reach the audit log, but only approval allows the sensitive operation to run. This is how the workflow places a human decision between a proposed action and its effect.
Policy determines whether execution needs human approval
Read the diagram as text
- Check policy. Evaluate the operation after validating the gathered evidence.
- Wait for human approval. Required for write or sensitive operations.
- Execute. Run the read-only operation or the approved sensitive operation.
- Record in audit log. Preserve executed actions and rejected decisions.
- Check policy → Execute: Read-only.
- Check policy → Wait for human approval: Write or sensitive.
- Wait for human approval → Execute: Approved.
- Wait for human approval → Record in audit log: Rejected; do not execute.
- Execute → Record in audit log: Record execution.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Threads, runs and checkpoints preserve different parts of state
Memory management saves and restores workflow state over time. A thread identifies the continuing conversation. The first question creates a run ID, and execution at a graph node saves a checkpoint. A follow-up stays on the same thread but creates another run ID, with checkpoints saved as that run reaches its nodes. Keeping those identifiers distinct lets the system connect multiple executions to one conversation without treating them as one undifferentiated attempt.
The local demonstration uses SQLite; Cloud Spanner is the named production option on GCP. Initially, the queried database contains no data. After submitting a question, a query returns a record containing the thread ID, the question and its execution time. Further queries inspect how many messages belong to that thread. The observable change is a persisted record of the interaction, rather than state held only in the running process.
What can an operator reconstruct from those records? The subsequent queries inspect thread and run IDs, the number of tools involved, the particular MCP tool executed and the progression through LangGraph nodes. Together, those views connect a user interaction to its workflow execution. The demonstration inspects stored state and execution history; it does not stage a crash-and-recovery test.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Store the run’s behavior before asking whether it paid off
Agent-level tracking records the submitted question, response, model, token consumption, cost and evaluation information. Gaggar recommends collecting it from day one. In his example, 100 agents have 100 projects with their own tracking and evaluation tables; those records can later be aggregated into dashboards. The purpose is to make questions about performance and return on investment answerable six months later, rather than attempt to reconstruct missing history.
The demonstration starts with six tables. After a query is submitted, the first inspection finds no data yet; a later refresh reveals the tracking row. What identifies that run and its result? The record includes request and session IDs, agent name, question, output, completion status, elapsed time, model and metadata. The delay between submission and visibility is part of the example: an immediately empty result does not establish that tracking failed.
The remaining views answer different questions about that same run:
- Quality checks: eight checks per run, with separate rows for dimensions including success, latency, cost, tool accuracy and safety.
- Evaluation scores: a single scoreboard row collects quality dimensions scored between zero and one.
- Quality metrics: a wide row uses nested fields for response, quality, safety, latency, cost and tool statistics.
- Tool invocations: each call records which MCP tool ran, how long it took and whether it succeeded.
- Agent trajectory: ordered records describe LLM calls and graph states, such as planning and summarization.
These tables establish a structure for inspection, not a validated quality score for the system. The walkthrough does not explain how the quality checks are computed or calibrated. Likewise, aggregated token and cost records can support an ROI investigation, but the demonstration does not measure business ROI. The useful foundation is the ability to relate a result to its model calls, tools, timing and evaluation fields.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Filter sensitive input before it spreads into storage and traces
A document upload may contain sensitive information, and a typed query may accidentally include an API key. The PII trimmer classifies data types and assigns severity levels. Gaggar describes blocking and redacting secrets when severity exceeds approximately 90%. That threshold belongs to the filtering rule; it is not a reported detection accuracy.
The intended interception point is ingestion, before the input reaches the model, memory database, tracking and evaluation tables or tracing system. This placement matters because preventing a model call alone would still leave a secret exposed if logging or checkpointing had already copied it elsewhere. The design aims to stop that spread early. Its effectiveness still depends on detection and redaction coverage; the walkthrough does not establish that every sensitive input will be caught.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Reduce the context before paying to process it
Large uploads and accumulated data can make each query expensive. The context optimizer shapes and prunes the payload before it reaches the model. It selects material within a token limit, reranks it for relevance, removes duplicates and compresses the remaining payload. Each operation targets a different source of excess: too much material, poorly prioritized material, repeated material and an unnecessarily large representation.
The first demonstration asks the agent to find all artifacts in JFrog Artifactory. The tracking view begins empty; after submission and refresh, a record reports 24,000 tokens used. Gaggar then enables the optimizer behind the scenes and reruns the same question.
How much does the recorded usage change? The optimized run reports 2,400 tokens, down from 24,000—a 90% reduction. The comparison makes the cost mechanism concrete: the question remains the same while the optimizer changes the material sent onward. The demonstration does not include a paired answer-quality study, so the smaller token count alone cannot establish that the answer retained all the required information.
ts ctx_opt_on input_tokens total_tokens ctx_saved
2026-08-27 05:40:36.692114 UTC true 2460 2958 23106
2026-08-27 05:39:14.895019 UTC false 24588 25130 null
A second example runs through a Slackbot and asks for Docker images across all repositories in Artifactory. Its first recorded run also uses about 24,000 tokens. After enabling the optimizer and repeating the question, the reported usage falls to approximately 1,500 tokens. Gaggar reports savings of 70–78% in most cases, distinguishing those broader observations from the larger reductions in the two demonstrations. Across many agents issuing many daily queries, reducing repeated context could materially reduce token consumption.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Make the supporting framework part of the first deployment
The closing recommendation returns to the portal’s original purpose: include the supporting frameworks before the agent ships. Every tool request, whether it originates locally or from a deployed agent, should pass through centralized authentication. MCP servers should control which tools can execute the requested operation. Monitoring and audit records should cover those executions, and high-risk actions should retain human approval.
Context optimization belongs in that initial framework too, because cost becomes harder to ignore as usage grows. The paved path makes these concerns repeatable: identity and resources arrive with the deployment; tool access follows a common route; decisions and executions leave records; context is reduced before model use. The engineering decision is to make those capabilities part of creating an agent, rather than rediscover their absence after deployment.
Suggest correction
This note stays in this page until you copy or download it. Nothing is submitted; reloading clears the draft.
Resources
Related talks
- Beating RL With Reflection: GEPA and Optimize Anything
Extends the evaluation discussion by showing how execution traces and human annotations can guide agent changes and train an evaluator.
Read the complete timestamped transcript
- 0:03
Hello, everyone. Um, I would like to walk you through that how we can transition an AI agents from an experiments to the real enterprise environment. So before I begin, a quick introduction about me. My name is Harish Gaggar. I'm a senior staff software engineer at Intuit Credit Karma, where I built and own the AI analytics infrastructure and services. Um, and beside that, um, I'm advisor, uh, author, researchers, um, and also a program committee members for various organizations where
- 0:33
I mentor the young researchers as well. All right. So let's, uh, walk through the, um, agenda here. So first we would like to talk about what are the challenges we have when we build an AI agents and deploy into the production systems, and how we, uh, can build a build different framework to address that, uh, different challenges here. Um, then how the architecture looks like for it. Um, then, um, how you can build your own MCP servers, um, and
- 1:03
MCP gateway, and then some monitoring on top of that gateway as well. Um, then further, we're gonna talk about what are the different type of agent framework you can have, um, and then, um, some additional framework around managing the memories, tracking and eval, um, how you can block the PII data, um, going before the LLMs, and then context optimizer to save your token, right? So without delay, let's, um, let me walk you through the challenge. Um, of course, like, there are
- 1:33
many agents, uh, uh, generally get failed in productions because their foundation controls are missing. Um, and the key reason behind is that, um, you know, like, oh, most of the people are missing the gateway to make sure that agent can, you know, communicate to the gateway to get authenticated first with the right auth providers so that they can redirect to the right MCP servers, right? Second, that, um, MCP servers are missing the right tool to execute your query. Third, do we have, like,
- 2:03
enough tracking and evaluation system to make sure that every query when you are sending to the agents, whether they've been tracked or not, and are you evaluating their quality response or not, right? Do we have any in-memory management system or not in terms of, let's say, you are doing the interaction with agent for long or for many days, uh, is it able to recall that what you have been asked in the past, right? And then there are other framework, of course, like do you have a context optimizers to save the token or do you have inner production in place to make sure that
- 2:32
your any secrets should not go into LLM or tracking or tracing or any memory DB or so, right? So lot of, lot of open questions. Um, so in order to resolve this, uh, we are not going to, you know, like, build a magical agent here in this demo. Uh, we're gonna talk about, like, different framework that you can build to address this. So first is the agent framework. Um, this is where, like, we use the, like, we choose to use the LangGraph framework, which is provided by LangChain
- 3:02
community. Uh, this is help us in building the task-specific node based on your questions, and which is easy to plug in any, like, uh, marketplace plugins or skills and prompts to make sure that agents, um, respond efficiently based on your query. Further, um, the second part is we are also building the MCP gateway to make sure that whenever agent make the call or any service account from that agent make the call, make sure that they are securely get authenticated to their
- 3:32
MCP gateway, um, and then get redirected to the right MCP servers. Third is the customized MCP, which means that you are building your own MCP servers with the right tool in place to execute your query. And then finally, some audit, uh, on top of what ex-actually operation you're performing, right? And then we talk about some additional framework around, um, you know, like how you can block the PII secrets, how you're gonna store the tracking eval into the cloud, um, and how you're going to persist the memories,
- 4:02
especially, um, any threads and checkpoints on every context. And then also how you can optimize overall your context to make sure that you don't bunch, don't burn too many tokens, uh, when you run your query. All right. So we'll talk about now the agent infrastructure. So the, firstly, we'll go through the local development. Um, the, the way the approach should be is that, um, you know, like, whenever you build an agent, right, first you have to register your agent into some
- 4:32
system. You usually tell that, "Hey, look, this is my, the use case, and these are the use case going to use for this particular vertical or system." Um, and you feed all those information and generate a ID for your agent, right? Once you have that, then you can have in dev portal, internal portal where you can ingest all the details saying that, "Hey, look, this is my agent use case. This is my agent ID. This is the framework I'm gonna use, either LangGraph, LangChain or so. These are my language I'm gonna use, either Python or TypeScript."
- 5:02
Um, and then you can choose, um, you know, like, your skills or prompts, um, and then you submit your informations, right, to this. Once you submit this, the dev portal is going to create a repository which is going to contains not just the agent framework, like your LangGraph agent framework, but is also going to contains, um, your memory framework, your tracking eval framework, right? And then your context optimizer framework and your PII blocking envir-- uh, framework as well. So the reason
- 5:32
for that is because not necessarily all your developers are taking care of all this framework, right? They just sometime, um, you know, create the framework what generally, like, AI assistant tool are generally support, right? Or they stick to one or two framework, right? So we just wanna make sure that when you roll out such framework, make sure they are consistent across the company, um, and they can take care of, um, and they can be centralized in one place so you can visualize the results as well, right?
- 6:02
So once you have that repository, so developers basically, you know, clone that repository, they can build their agent as a Docker image into containers And test their agent input and output through the query, and they can also test this on their local memory, maybe Cloud Spanner Evaluator here. And you can also store this tracking eval in some of your DB, maybe SQLite or maybe BigQuery, um, to see, to make sure that agents are fully working on your local, right? So once your agents are running in local, um, then you're ready to go ahead and generate a
- 6:32
PR, and then you can, once you get your two PR approval, then it's go to the CI/CD pipeline, right? So the PR is basically going to, firstly, is going to, uh, register your agent into some centralized agent registry system, which provide a Langfuse URL to it to make sure that you see all the traces. And then the CI/CD pipeline is basically, you know, whatever the local Docker container you built here is going to build that container into production system. Uh, it will scan that image, make sure that there's no vulnerability. Um, if everything looks great,
- 7:02
security approved, you can move that e-image back into the registry, right? Further, you have your own deployment platform or orchestrator platform, which going to deploy that agent into the Kubernetes cluster in a given namespace, right? So the good part here is that this dev portal is not just creating the repository, um, right? It also do create, um, bunch of other production resources. For example, the workload identity, um, right, and then memory DB or
- 7:32
tracking and eval tables, and then agent secret management, right? Um, so it based on whatever the name we provide, it not just, again, like, not just repository, it create all these different resources as well. So what happen is when this internal orchestrator, when it deploy the agents, the agent already know that, "Hey, what workload identity I supposed to work with," right? That binding should already happen. So once this case to GSA binding happens, the agent will start communicating to all these resources.
- 8:02
Once everything in place, um, agent is further going to communicate to rest of your endpoint services like for example, AI gateway or MCP servers or Langfuse for tracing or some other API endpoints, right? So now firstly, we talk about the MCP server as well. So we talk about architecture, now we're going to the different component. So MCP server is something that you can definitely use something which is outside available. But if you're working on air gap environment, it's highly recommended that you build your own
- 8:32
MCP servers. So, um, yeah, uh, and you can use the, um, I will not go with the definition, everyone knows about it. So the way the framework has been built here is that w-you, you can leverage the FastMCPServer package, um, or wrapper to, um, you know, build that MCP servers, and you can create a different set of tools inside that FastMCP wrapper to construct your, um, you know, the overall servers. And then you make sure that, you know, like every tool when you
- 9:02
define an MCP server, they had to be fully reviewed by your own internal security team to make sure that only authorized tool can be go in production. And that's how you can, you know, enforce some filters, limits and block the fields, which is really not required to go to LLM and get you the information which is you're not supposed to get, right? Um, and third part is like a, of course, you need to make sure that you build your MCP server in a way that whenever the traffic is
- 9:32
coming from your MCP client to MCP gateway to MCP servers, either is coming from user or service account, there has to be, uh, authenticated to the MCP gateway first, and so that they can be access your MCP tool, um, to get that right information. All right. Um, now the gateway part. So the how the gateway works, so this is again, you can build it your own internal MCP gateway if you are again working on your air gap environment. So the way it work is you are working on Claude Assist- uh, like local,
- 10:02
you know, code assistance tool or maybe your agent, right? So the moment you submit your first query to your agent, it will send that one-time authentication request to the MCP gateway, right? MCP gateway recognized, uh, and take your request. First it will save your session state into the Redis, right? And then it will, um, you know, return back the JWT access token back to the, your client, right? Once client will receive that JWT token, um, it will know that what MCP server needs to call, right? So it
- 10:32
will submit that tool/call request on a given namespace/MCP, meaning, let's say, artifactory or githubs/MCP with the auth bearer, uh, JWT, um, token, um, into that. So once MCP gateway, um, receive that, uh, it will understand that what namespace, meaning MCP server needs to look up for, and then it will unwrap that JWT token, and then it will attach that user token in that request and send it to the respective MCP servers. Now, once MCP
- 11:02
server receive it, it will make that respective API calls to that endpoint of that, um, you know, uh, to the service, and then it will exchange that ID token, uh, and return the response based on whatever scope that user is authorized to see, right? So once the user receive the token, um, you know, uh, MCP server basically, you know, unwrap that all the data, make the payload which is compatible for MCP gateway to receive. And then finally, the final payload will send back to the client to see the data
- 11:32
points. So what it shows is that, you know, like we build this MCP gateway to make sure that we have a centralized authentication and routing hub that securely connect any coding assistant tool or agents to the endpoint via this MCP gateway and MCP server. There, there should not be direct access provided, um, on any endpoints. There has to go through the MCP gateway through proper authentications, uh, before they're reaching to the MCP server and the endpoints. All right.
- 12:03
So now the part is the gateway monitoring, right? We talk about, again, we talk about architecture, we talk about MCP server, MCP gateway. Now it's a monitoring. Again, the AI agent cannot be trusted without continuous monitoring, evaluation and feedback, right? This is why we are building the MCP gateway s-uh, monitoring. So the way the monitoring work is usually, like, the moment you submit any calls to your agent- Uh, it will send all the operation log into the log store. You can use ELK, Loki or CloudWatch also, right?
- 12:33
And generally, this contains information like show me like all last photo ones in my last 10 minutes call, right? And what did or do when the token refresh failed, right? So those kind of requests can go to the log store. Now, there are some, you know, traces and metrics data, um, like real-time data that goes to the OpenTelemetry as well. Um, OpenTelemetry again is open source solutions, which going to collect all your traces, um, and metrics, and is going to process the data and
- 13:02
redirect to the back-end system, right? And in this case, the traces will go to the Tempo. This is again a open source solutions, which going to store all your traces in terms of like, uh, traces in the sense like, you know, like information around like which tool was slow on this request, whether this call is coming from agent or IDE, uh, what was the exact error, um, on this failed call or so, right? So these kind of data are going to store into the Tempo. Oh, my bad. Okay. And then the metrics is about,
- 13:32
uh, the Prometheus is all about the information about like, um, what are P95 latencies for a given tool, right? Um, whether my auth, uh, rejection is spiking on a particular time or not, right? So those kind of metrics is going to store in Prometheus. So once you get all this data point, like traces and metrics, this entire aggregate data back actually save into the Grafana for your visualizations. Right. All right. So
- 14:03
the MCP server gateway and monitoring is covered. Now we're going to talk about the, uh, framework. So the agent framework, right? So we use LangGraph, uh, m- for most of the cases. Um, and it's basically LangGraph, again, is, uh, uh, provided by LangChain community, is provide a graph-based architecture that support, uh, loops, branching, state management for complex multi-agent AI workflows, right? And the way it work is like the moment you submit your first
- 14:33
query, uh, we do interpret, um, you know, like your question first, like what type of questions it is and what response it's looking for. Um, and based on that, uh, it's create a plan, right? Um, then it will send for classifi- uh, classify, uh, area where it will determine that, hey, whether it is just like a read operations or whether it is just like a, um, you know, modification level changes coming up or whether you really want to delete something, right? Or some sensitive operations.
- 15:02
Based on that, um, it adjusts the overall structure and then plan accordingly. The next step is plan, is actually itself is a subgraph where you do the major of your operations. Meaning that in the plan, like once you know what operation you're performing, it goes to the, like, let's say in the fetch evidence section, where it actually call that whatever the question you ask here, after all interpret and classify, um, it actually go ahead and make the call to the MCP gateway, uh, make
- 15:32
sure they get authenticated, um, and then it make the call to the MCP server using the right tool, um, and get that data, and then generate that evidence, right? Once you validate all your finding, then it will check the policies. Uh, it will check whether it is a read-only, then go ahead and execute, and w- I will save all the record points into our audit logs. But if it is a write and sensitive operations, definitely it will go through the HITL process where you wait for an approval. If it is approved, you go ahead and execute. If it is
- 16:02
get rejected, in either case, it's going to save into the audit log, right? So if we combine all this together, like MCP gateway, MCP server, um, and the agent, right? This is how things looks like. So ultimately, whatever the agent you are building, whatever the call you are submitting, it will go through the MCP gateway for authentications, and then it will redirect to the right MCP servers to execute, uh, you know, your query on a given
- 16:32
tool, uh, so that it make the API call to that endpoint and get the response, right? All right. If you t- if you think from the infrastructure perspective, this is how it looks like. So agent is communicating to the MCP gateway, goes to the MCP server and to the endpoint. All right? So now the, uh, we talked a lot about MCP gateway, MCP servers, um, the monitoring, um, and the agent framework. Now we'll talk about, uh, three and four additional
- 17:02
framework that we should have to take care of when building the end-to-end ecosystem, right? One is them is agent memory management. Again, it's, uh, uh, you know, it's a, it's refer-- it's basically a process where you can save and restore the state of your workflow over the time, right? Ensuring that, uh, it's provide you a durability and seamless recovery across, uh, across your every query run. So the way it work is like whenever you submit any query to your agent, um, it
- 17:32
actually generate a thread, right? Uh, on that given thread, uh, when you ask your first question, it will going to create a run ID for that. So the moment you ask some deeper questions, it will go on a particular node in that LangGraph and execute that query. And on that node, it going to save that respective checkpoint here. Right? Now, if you ask a follow-up questions on the same thread, it going to create another run ID. And again, depend on your follow-up question, it go to a specific node
- 18:02
and save the checkpoint over there, right? And all this data for your local development, you can use the SQLite, SQLite DB, uh, but in production, unless in GCP, you can use, um, Cloud Spanner and, um, in respect to databases, you can use in Azure, Azure and AWS as well, right? So just walk you through the quick demo here, um, on the memory. So for example, um, I have, you know, like just build this agent locally and then, um, you know,
- 18:32
trying to run this query on SQLite DB to see if there is any data or not. Um, so I'm just You know, initially I don't have any data. Now I submit my query, you will see that I receive the one record, which is saying that, "Hey, this is my thread ID, this is the question I have been asked, and exact time, um, you know, when this query was executed." Now, if I run the next query to see that, "Hey, how many messages have been produced on a given thread?" Um, you can get this information by running the further queries, right? Now, if I would
- 19:02
like to run to see that, um, you know, like all the metadata level information on my query, um, I can run my query again, and you can see the detailed description around, you know, like how many tool I get involved, what is my thread ID, what is my run ID, um, you know, like, and all those informations. Now, if I want to also see that, "Hey, what tool have been executed when I run this query?" Um, you can go ahead and see the exact tool name as well on that given MCP servers. Now, further I run-- a-also want to see that
- 19:32
how the progression was happened on my LangGraph framework. That also you can see by, uh, running this type of query on your me-memory DB. You can see that here, like different kind of, you know, nodes been called when actually, um, you know, that query has been run. All right. Now, the tracking and eval. So just to make sure that, uh, there is no confusion here, like, uh, we talked about earlier the MCP gateway monitoring, this is like agent level tracking. So we do this
- 20:02
because we really want to track that what query you have been submitting to the agent, um, and what the response is coming, what LLM model you have been used, what's the cost of your token, uh, how many token you have been burned, and what's your, you know, evaluation criteria looks like. What is the quality of your response or not, right? So you need to store that data from the day one. And imagine that if you have a thousand plus agent, every agents will have their own, um, you
- 20:32
know, um, agent tracking and eval, um, project and the tables, right? So if you combine, let's say if you have hundred agents, you will have hundred different projects, and each project has their own respective tracking and eval tables. Now, tomorrow, six months down the line, if somebody asks you that how your agent has been performed, what was the response, what was the quality, how much token you've been burned, you can aggregate all this data and you can create executive dashboards around it to see, um, whe- to see the what is the ROI in building so many agents,
- 21:02
right? So let me walk you through the, the tracking and eval, uh, demo here. So, um, here, like, if I look in, uh, this one, so first, like, you know, like if you see that there are six tables are listed here. One is a job tracking table, um, and which is contains most of this, like, information around what query, like what is your query, what's the response and everything. And there are a bunch of other tables as well, like your quality related and your tool
- 21:32
invocations and then node level executions and all that. So in this case, I'm going to submit my first query on my agent, um, and you will see that, um, there will not be, uh, any data in the first place, uh, because it takes some time, so I will refresh this again. And this time I will see that when I refresh, I do see that for a given query, I get my request ID, my session ID, my agent name, um, you know, what question I've been asked, um, and what was the output,
- 22:02
um, of that, uh, question, uh, whether it has completed or not, and then what is the time it took, what is the model I've been used, um, and then some of the meta le-le, uh, metadata level informations, right? Further, if I look for the quality, again, there is a eight quality checks per run, right? And you can see that this one row each for, you know, like success, like latency cost, tool accuracy, safety, et cetera, like all this has been stored here, right?
- 22:34
Now, the eval scores. This is again, this is going to be a one row. This is going to be a scoreboard for your given run, which is basically varies between zero to one, which is scores, uh, you know, like all the quality dimension in a single row, right? Now, the quality metrics. So again, it's a deep dive on that particular one run. Basically a nested field for the response, quality, safety, latency, cost, uh, and the tool stat in one wide row, right? Now, the tool
- 23:03
invocation. So every tool call, which MCP tool ran, how long it took, and whether it succeeded or not, right? And now the tool traj-- agent trajectory, which is talk about step-by-step story of the run, each LLM call, and the graph state in order, like plan, summarize, et cetera. All right. So now, uh, the next is the agent PII trimmer. This is very, very important because many times happen
- 23:33
that when you're working in agents or, uh, this thing, um, sometime there is a possibility that there are some secret being ingested as, as a part of your query, right? Let's say you are uploading some doc which has some sens-sensitive informations, um, and sometime while typing, you may have some API keys or so. So we have a framework which define that different sort of data types and their severity level. If your severity is above ninety percent or so, uh, we will block and redact that, uh,
- 24:03
secrets, uh, immediately so that, uh, make sure that that secret is never goes to LLM models, right? Neither it will goes to your memory DB or nor it will go to tracking and eval tables or tracing systems as well. So the goal here is to just block this sensitive data as it is the moment it get ingested, right? Moment, moment you type. All right. Um, the last framework is all about now is the context optimizer. This is one of
- 24:33
my favorite area where like, um, you know, uh, whenever you submit... It's very important, right? Whenever you submit too many queries, sometime you upload lot of the um, data points and you upload bunch of files systems or so, right? It's very important that time is, um, you know, like you make sure that you save your token in every query. So we build in context optimizer framework where you actually shape your query, and then you prune it before you go to LLM. In the pruning system,
- 25:03
uh, we do select the payload based on that token limit, then we re-rank that, "Hey, which is the most relevant, relevant information that we should target for when you submit this huge giant, you know, context?" And then we dedupe that, uh, remove all the duplicates and create a very unique set of, uh, very curated data like, uh, I mean, the main data for the LLM. And then we finally compress this large payload before it goes to the LLM. So let's see this, uh, demo here.
- 25:33
Uh, this will be quick. So for example, let's say I will submit my first query, "Hey, find me all the artifacts in my JFrog Artifactory." Um, you see that first when bef- even I not submitted any query yet, there is no data here. So the moment I submit this first query,
- 25:51
and if I refresh my query here, um, and then you can see that the record has been, uh, added. So you can see there is a twenty-four thousand token has been used. I will go back a little bit, um, just to show the thing. So you can see this twenty-four thousand token has been used for given questions. Now if I ask, uh, now I enable the context optimizer behind the scene, and I'm rerun the exact same query again. So this time you will see that, uh, you know, for the same query, if I refresh this run,
- 26:22
and you will see that, uh, this time it used just two thousand four hundred token, right? For exact same query. So which is almost like ninety percent saving. I can rerun the same demo on the Slackbot as well, where I'm asking Slackbot, "Hey, tell me all my Docker images across all the repositories in Artifactory." When I run this query, there is no data. Now I'm going to submit my query. So once this query will run, um, and if I run my query again here, and you
- 26:52
will see the record get created, it's again used like twenty-four thousand for this, this particular query. Now, what I will do is I will, uh, turn on my optimizer behind the scene, and I rerun the same query in the Slackbot. Um, and then if I run this query again, um, it will show like it exactly used one thousand five hundred almost token, which is again like ninety percent saving. In most of the cases, we get like seventy to seventy-eight percent token saves. So imagine that you have thousands of, um, you know, like,
- 27:21
um, um, if you have thousands of like agents which is running like, you know, many queries per day, imagine how much token you can save with this framework. All right, so now the key takeaway here is, uh, you know, like, um, for us is like make sure that we build an enterprise framework, um, not just an agent, right? Which makes sure that you have all the framework underneath. Don't try to be in rush to building some agent quickly and deploy, ship it. Make sure that it has to go through
- 27:52
all the framework altogether, so that you should not realize at a later stage that, you know, I should have this framework in place, right? I-I-In the very first place. Second is like your make sure that you have a proper gateway system. Make sure that every tool call request, either you are doing this from your local or through agent, it has to go through some sort of centralized MCP gateway. Uh, make sure it authenticated before it reaching to the MCP servers. And then make sure that your MCP servers take a control
- 28:21
on what tool, uh, you know, uh, y-you should use, uh, to execute a particular query, right? And make sure that you have monitoring in place and complete auditing in place for that tool executions. And of course, you should, uh, you know, keep the always human-in-the-loop process for every, um, every action items, every high-risk items. Uh, make sure that there is a observability, um, and, you know, um, auditing has to be in place. And last but not least, make sure that context optimizer should
- 28:51
always be there to justify your cost, um, um, on your agents, right? Um, so just want to share is that I released my, um, MCP, uh, book recently on Amazon. So feel free to download, it's free in Kindle, um, and it's, uh, hard copy is also available. All right. Um, thank you so much for watching this video. Um, and if you do have any questions, feel free to, you know, like, reach out to me over LinkedIn
- 29:21
or Twitter. Uh, I'm happy to answer. Thank you again.